occur in injured tissues, these must be under some type of control by chemical agents having hormonelike qualities. The whole question of the reason why cells derived from a single fertilized ovum develop differently to become nerve, muscle, bone, or blood cells and so on is one that suggests that specific chemical inciters are involved. The nature of these hypothetical agents is unknown. These are fields in which a concentration of effort is urgently to be desired. Our understanding of the dedifferentiation and the aggressive growth of malignant cells waits upon the elucidation of the chemical factors that control normal differentiation and the orderly growth of healthy organs.

THE CHEMOTHERAPY OF CANCER

Until the turn of the century laboratory research on cancer was essentially morphological. That is to say, it was preoccupied with detailed examinations of the anatomical structures, both macroscopic and microscopic, of all sorts and varieties of tumors. Knowledge of the functional aspects of the growth of cancers was meager depending, as it did, on the accumulation of such desultory observations on human subjects as alert clinicians were able to make. No experimental study of the growth of malignant cells was possible until actively growing tumors could be brought into the laboratory and studied under controlled conditions.

In the first two decades of the century 3 dramatic developments followed one another in rapid succession and laid the groundwork for an experimental science of cancer. The first of these was the demonstration that spontaneous cancers in laboratory animals could be transmitted to other animals of the same species by the implantation of a few of the cancerous cells. The tumors that developed retained the essential characteristics of the original cancer and could, in their turn, be used to infect other animals. In this way the supply of a given type of tumor could be maintained indefinitely. Somewhat later it was shown that cancers could be produced in the laboratory by simple chemical means. All that was necessary was the repeated application of coal tar to the skin of animals. The third development, was the introduction of the technique of tissue culture. This was immediately applied to cancerous tissues and it was shown that cancer cells could be grown in test tubes and would retain their malignant characters indefinitely.

Within a few years these new techniques had been developed and refined to such a degree that a wide variety of standard animal tumors became available for such studies as biochemists, geneticists, virologists, and endocrinologists might be inspired to undertake. The stage was also set for a systematic approach to the chemotherapy of cancer. It became possible to compare the growth-inhibiting potencies of an indefinite number of drugs on a variety of animal tumors and so to seek out those classes of compounds that showed promise of having therapeutic value. In this way the arduous and exacting business of testing drugs on human cancers could be limited to those compounds that survived the screening process in animals.

Although, technically, the stage was set some 20 years ago for an intensive chemotherapeutic program, the guiding ideas and the impetus to develop it were lacking. Only sporadic studies by a few enthusiasts developed. The objective of such a program was clear. It was to seek out substances that would destroy cancer cells without doing irreparable damage to the normal tissues of the host. What was lacking was some guiding principles as to where to look for these materials and a sustained faith that they could be found. The lack of faith seemed to be justified by the indifferent success of chemotherapy in the broad field of infectious diseases. If it was proving so difficult to find drugs that would discriminate between cells as utterly different genetically as bacteria and the cells of the human body, what hope was there of finding agents that would leave normal cells unharmed while they sought out and destroyed the cancerous progeny of these cells? It was recognized that tumors occasionally did regress spontaneously and so were presumably subject to some indigenous inhibitors of growth. On the other hand, so much emphasis was placed on the principle that malignant growths were autonomous that there was little encouragement to attempt to control them pharmacologically.

The discovery of the sulfa drugs, followed swiftly by the discovery of the antibiotics, changed the whole climate of opinion. It became evident that the world of bacteria could be brought under control. There remained the world of the viruses and beyond that the realm of the malignant cell. The task was still recognized to be a formidable one, but it was no longer to be rejected as

hopeless. The chemotherapy of cancer began to take form. It received impetus from the fact that our understanding of the chemistry of growth, of the nutritional requirements of the cell, and of the role of the vitamins in growth, had developed to the point that ways of controlling growth by chemical agents were beginning to suggest themselves. The most significant suggestion actually came from the study of the mode of action of the growth-inhibiting properties of the sulfa-drugs. It was found that these drugs were interfering with the utilization by the bacteria of a vitamin (para-aminobenzoic acid) that was essential for their growth. Because the sulfa-compounds were related in chemical structure to the vitamin it was postulated that they entered into competition with it and blocked its utilization by the cell. The drugs behaved like fraudulent vitamins, deceiving the cell to the extent of being incorporated into the metabolic machinery in place of the true vitamin. Since, however, they were imperfect substitutes, they were unable to carry out the metabolic functions of the vitamin. The gears of metabolism became locked, growth was arrested and the cell died.

This hypothesis has come to be known as the antimetabolite theory of the action of drugs. It has proved to be a very profitable one in explaining the action of many substances foreign to the body. It has provided chemotherapy with its most useful guiding principle, namely, that if one wishes to find a drug that will inhibit a specific metabolic reaction one should look for it amongst the chemical analogues of the metabolites that normally enter into the reaction. This is exactly the kind of principle that the synthetic organic chemist needs as he attempts to keep the biologist supplied with new compounds worthy of chemotherapeutic evaluation.

It is probable that research on the chemotherapy of cancer would have gotten well underway in the early forties had it not been delayed by the outbreak of the war. Paradoxically, the first dramatic success, the first breach in the chemical defenses of the cancer cell, came as a byproduct of wartime research. One of the early symptoms of exposure to nitrogen mustard gas was found to be a fall in the count of white cells in the blood. Since normal white cells were destroyed it was natural to hope that the overproduction of malignant white cells that is characteristic of leukemia might be brought under control by the same agent. The experiment was tried and was a success. Hundreds of related compounds were promptly prepared and tested on animals but, curiously enough, the original war gas was found to be about as effective as any. It came into use therapeutically and proved to be a helpful palliative in some cancers of the blood-forming organs. More recently a distantly related compound, triethylene melamine, has come into favor.

It is interesting to note that the two discoveries that gave to the chemotherapy of cancer the impetus which has been so evident since the end of the war arose from investigations unrelated to cancer. The antimetabolite theory had its origin in studies of the nutritional requirements of micro-organisms. The therapeutic potentialities of mustard gas were disclosed in the course of research in chemical warfare.

It has been the recurrent theme of this report, that, because cancer cells multiply more rapidly than do most normal cells, they may be more vulnerable to nuclear damage. The manner in which the nitrogen mustards exert their lethal action supports this thesis since there is good evidence that their primary effect is on the nucleus of the cell. They probably interfere with the metabolism of nucleic acid. Indeed, they are one of the few classes of compounds that increase the rates of genetic mutations in cells to about the same degree as do X-rays.

In line with current thought, contemporary programs directed toward the chemotherapeutic control of cancer are now sharply focused on the nucleus of the cell. To begin with, drugs such as colchicine, podophyllin and urethane, that were known to inhibit cell division were screened on animal tumors. In general, the results were disappointing. The differences between the doses toxic to the animal and the doses that inhibited tumor growth were too small to encourage therapeutic use. Only urethane has received any measure of clinical acceptance. It is stated to be helpful in the control of multiple myeloma, a condition that has not responded to other cancerolytic agents.

Reaching out beyond the known drugs, extensive synthetic programs were developed to produce compounds that might be antimetabolites in the synthesis of nucleic acid. These have been of two main types. One type comprises a great variety of analogs of the two major constituents of nucleic acid, the purines and the pyrimidines. A few of these, such as 8-azaguanine and 2.6

diaminopurine have attracted particular attention and have received clinical trials. The second class of compounds are those which have come to be known as antifolics. This is because they are related to and competitively inhibit the vitamin called folic acid. When these compounds first attracted attention the relation of folic acid to growth was obscure. Recently, however, it has been shown that this vitamin is essential for the synthesis of nucleic acid and so the antifolics are to be regarded as inhibitors of nucleic acid metabolism. Several of these compounds, notably Aminopterin and A-methopterin have received favorable publicity. Their therapeutic value, however, is limited to cancers of the blood-forming organs and is only temporary. At doses tolerable to the host they do not damage the cancer sufficiently to prevent recurrence following removal of the drug. It is claimed that they are particularly useful in the control of acute lukemia in children where remissions are obtained in about 50 percent of cases and a second remission may often be induced before the cells become refractory to the drug.

It must be acknowledged that the attack on the nucleus of the cell has, as yet, resulted only in minor successes. The therapeutic value in man of the antinucleic acid agents has been confined to cancers of the blood-forming organs. Several important lessons have, however, been learned and will guide further studies. In the first place each type of tumor in animals shows its own range of sensitivities to a series of drugs. Moreover, the same type of tumor in different species may show differences in relative sensitivities. This means that each type of tumor in man represents a separate chemotherapeutic problem. It is unlikely that a single miracle drug will be found which will control the major types of cancer that afflict mankind.

This individuality in response that is shown by different kinds of cancers greatly complicates the problem of screening compounds since there is no clear relation between the responses of animal tumors and the therapeutic effectiveness in man. If human cancers could be transplanted to animals a more reliable screening of compounds might be possible. Unfortunately when the transplants are made in the usual ways they do not survive. Recent work using novel methods of transplantation give promise of developing procedures whereby actively growing human tumors may be cultivated in animals.

A second lesson that has been learned has been the dismaying ease with which tumors develop resistance to a particular drug. In this respect they are imitating the well-known adaptability of bacteria and other simple forms of life to toxic agents. It is probable that the cells in a single tumor vary widely in sensitivity to a given drug. A dose that will cause a notable regression in the size of the tumor does not kill all of the cells. The hardy ones survive and a new tumor develops populated by resistant cells with the result that a second dose of the compound is less effective. The ideal treatment would be one which killed every cell in the cancer. This would require a very much higher dose which would probably seriously injure the host if it did not kill him. This is an inevitable limitation to the antimetabolite method of attack if the metabolic process which one seeks to block is essential to the growth of normal cells as well as to that of malignant cells. One is attempting to discriminate on the basis of a quantitative rather than a qualitative difference in sensitivity. It is to be expected that the sensitivity of some normal cells will actually be greater than that of the most resistant of the cells of the cancer. In this situation the dose that will kill all of the malignant cells must destroy some normal cells.

One escape from this dilemma which has been suggested is the use of combined therapies. A priori, it is reasonable to assume that the cells that are abnormally resistant to one drug will not be the same cells as those that are abnormally resistant to a second drug that has a different mode of action. The simultaneous use of the two drugs should, then, result in more damage to the cancer at less hazard to the host. An alternative approach would be to seek out some qualitative difference between normal and malignant cells. If a metabolic reaction can be found which is essential to the growth of cancer cells but is not essential to normal cells it should be possible to block this reaction completely with an antimetabolite without interfering with the metabolism of the cells of the host. The remarkable efficacy of the antibiotics may well depend on the fact that they inhibit reactions which are essential to bacteria but not to human cells. There are a few hints that qualitative as well as quantitative differences in the metabolism of normal and of malignant cells do exist but their nature is not yet clearly enough defined to be ready for chemotherapeutic exploitation. Progress along these lines waits upon fuller knowledge of the detailed pathways of metabolism.

It is a matter of definition whether biological materials should be classified as chemotherapeutic agents. In the case of the steroid hormones it seems quite appropriate to do so inasmuch as a number of synthetic steroids, such as stilbesterol and methyl testosterone, have therapeutic activities similar to the hormones of the sex glands. Some reference has already been made to the inhibiting effects of these substances on human cancers. The present status of hormone therapy may be summarized in the statement that it is largely confined to the control of tumors of the breast and of the prostate and that hormones have their greatest value when the diseases are beyond the control of the surgeon or the radiologist. In cancer of the breast in patients well beyond the menopause, estrogens have been effective in inducing regressions of the primary tumor and of its metastases to viscera, skin and lymph nodes. Prior to the menopause, on the other hand, androgens have sometimes been useful palliatives. Castration is often effective in cancer af the male breast. Castration also leads to regression of tumors of the prostate and of many of their metastases. Beneficial results are only temporary, relapse occurring after a year or so. Estrogens alone or coupled with castration have also been used in the control of cancer of the prostate.

An interesting aspect of hormone therapy is its flexibility. One may treat the patient either by depleting the body of its normal supply of hormone by the removal of the gland that secretes it or one may augment the patient's own secretory activity by the administration of additional hormone. Further than this, one may exploit the antagonism between the male and the female hormones by the administration of estrogens to males and of androgens to females.

Although it would be improper to classify a virus as a chemotherapeutic agent it is appropriate to conclude this review with a brief reference to the use of viruses to destroy cancers. Starting with the knowledge that viruses are highly specific with respect to the types of cells that they will attack, a search was begun for viruses that spontaneously showed or could be trained to develop an affinity for tumor cells. The virus that has shown the most ineresting results is one that was isolated from an epidemic of encephalitis in Siberia in 1939. This virus normally attacks the brain and is lethal to man. When it was screened on animals carrying different kinds of tumors it was found that the virus entered and multiplied in the cells of some of the tumors and destroyed them. It was highly selective in respect to the types of tumors that it would attack. By continuous passage of the virus from tumor to tumor in animals it has been possible to enhance its tumor-destroying ability. Unfortunately it has not been possible to reduce its affinity for nervous tissue. Other viruses that are not lethal to man are now under investigation. When one recalls the diversity and the mutability of viruses it is evident that a wide and fascinating field of investigation is being opened up. Proverbially it is good tactics to set a thief to catch a thief. It may yet prove to be sound therapy to set one aggressive growth to destroy another.

THE END OF THE BEGINNING

The review of current trends in cancer research to which the preceding pages have been devoted is necessarily superficial and incomplete. Many aspects of laboratory and clinical research have been ignored. In particular, there has been no attempt to describe the steady progress that continues to be made in the care of the cancer patient. Improved methods of diagnosis, refinements in surgery, the increasing range and power of radiotherapy, the better control of the use of anaesthetics, of transfusions and of antibiotics, the increasing concern for the physical and psychological rehabilitation of the cancer patient-all these are combining to raise the cure rates of one or another type of cancer by 1 to 2 percent a year. The cumulative effect is not to be ignored. Whatever revolutionary advances in the control of cancer may or may not develop in the next few years this kind of slow relentless progress will continue. It reflects the steady onward march of medical knowledge as a whole.

In his excellent book, Science versus Cancer, Dr. I. Berenblum has summarized his view of the current status of cancer research in the words delivered by Winston Churchill at a critical stage of the last war:

"Now this is not the end. It is not even the beginning of the end. But, it is, perhaps, the end of the beginning."

The quotation is apt. It accents the main argument of this review that the problem of cancer is beginning to take definitive shape. The ground has been cleared, there is a common understanding of objectives between a diversity of

scientific disciplines, the foundations for a comprehensive experimental attack on the causes and cure of cancer have been well laid, work is being pressed at an increasing tempo.

In the pages that follow there will be found some detail of the research and training programs that have been the responsibility of the committee on growth. Brief synopses are given of the nature of the individual research projects that are now in operation. The reader may find them bewildering in their diversity and forbidding in their brevity of description. He should, however, find in them some reflection of the major trends in cancer research that have been discussed above. One point of first importance bears emphasis. The ideas behind the individual projects, the ideas that gave them birth, did not come from the committee on growth. They came from the investigators who are conducting the projects. That is to say the research program of the committee is not a program conceived by a group of 20 advisers in executive session, it is the program of a congress of several hundred operating scientists.

In some respects it may be said that the policy of the committee on growth is one of laissez-faire. It is not, however, a random policy simply because the progress of science is not random. It has a strong sense of direction, intuitive but compelling. It is held to its course as though by an automatic pilot. By some mysterious telepathy, the thoughts of independent investigators mold each others emerging ideas and there develops a consensus of opinion that is the real architect of the pattern of research at any time. No committee, however wise, can substitute for this consensus. No wise committee would attempt to do so. Dr. CAMERON. Dr. Heller has pointed out that this is not one disease. It is tremendously important to understand how many diseases we are actually talking about when we use the word "cancer."

One of the great students of cancer not so long ago, Dr. James Ewing, used to say that in terms of their behavior and their structure, there were more kinds of cancer than there were of other diseases combined. There is the type of cancer which is similar to cancer of the breast, cancer of the rectum, and cancer of the lip, and these bear no relation to one another in terms of cause, in how they begin, in how they are treated, and how they respond to treatment.

If we take one kind of cancer, cancer of the breast, we still have an enormously complicated problem, because there is not 1 kind of cancer of the breast, but there are 6 or 8 general kinds, but if you were talking about 1 kind of cancer of the breast, what the physician calls adenocarcinoma, you still have a complicated problem facing you, because there are many kinds. They are roughly classified as four grades according to the rate at which they grow. Grade 1 does not behave like grade 4 and a patient with grade 1 follows no pattern that is comparable to the patient with grade 4 of the same type of cancer of the same organ.

Now, it is that inherent complexity of cancer which is responsible for the complexity on the whole which Dr. Heller has referred to. There is complexity of public education, complexity in diagnosing it, the complexity that faces the doctor who is trying to arrive at a reasonable validation of his suspicion of cancer and the complexity of treatment. Perhaps most important to us all here today is the complexity which this inherent field of cancer imposes upon research.

Fifty years ago the annual death rate from cancer in the United States was 64 per 100,000 of population. The estimated death rate for the current year is in the order of 145 per 100,000.

Mr. Chairman, if it would be consistent with my discussion of this problem I would like to suggest that you open the document which is before you called Cancer in the United States and turn to page 14, because it will hasten my presentation if I can refer to the charts which appear in that document serially.