Two decades ago, one scholar of cancer expressed the opinion that to find a chemical that would make cancer disappear and leave normal tissues unharmed was almost equivalent to finding a drug given by mouth that would dissolve one ear and leave the other in place. And, up to that time, with the research literature littered by negative reports and by tarnished reputations of investigators, this pessimism was perhaps justified.
Two discoveries, however, led to a revision of scientific opinion that stimulated a renewed interest in the chemotherapy of cancer. The first was an accidental observation during World War II, that a potent war gas, sulfur mustard, produced selective damage to the lymphatic system and the bone marrow. Under the cloak of wartime secrecy, a related compound, nitrogen mustard, was tried cautiously in patients with cancer of these tissues, leukemia, lymphosarcoma and Hodgkin’s disease. Remarkable temporary remissions of the diseases were achieved in a good proportion of such patients.
The other discovery was made through investigations that sought to block metabolic processes of cells by providing the cells with chemical relatives of essential materials that the cells could not distinguish but which did not fulfill their needs. These are the “antimetabolites.” One of the first series of compounds of this nature, the antifolic acids, were shown by Sidney Farber of Harvard to have beneficial effects in children with acute leukemia. Apparently the leukemic cells were affected to a greater extent than normal cells by the interference of metabolic reactions requiring folic acid.
These findings were used by Farber and by the late Cornelius Rhoads of the Sloan-Kettering Institute in New York as evidence for the need of expanded drug development and testing programs, and received sympathetic response in our Congress. By 1955 a large National program of cancer chemotherapy was organized in the United States, and is now in full swing. Over 100,000 chemicals and other materials have been tested on laboratory animals, and over too chemicals were deemed sufficiently effective and interesting to be tried in patients with cancer. The British workers entered the field with great force and achievement at the same time, and there are important chemotherapy programs in Germany, the Soviet Union and Japan.
Research in the chemotherapy of cancer can be divided into four main steps: the selection of the chemicals or other materials to be tested; the bioassay of the material on some animal tumor system; preclinical pharmacology of agents that are selected on the basis of antitumor effects in animals; and, finally, clinical trials.
The selection of agents to be tested may be made in several ways. Of course, the neatest method would be to start with a clear hypothesis, and plan for the specific chemicals by which the hypothesis can be tested. This is the “rational” approach. The extreme opposite of this is blind empiricism, in which any chemical is swept off the shelf and introduced for animal testing. In most instances, the choice is based on considerations somewhere between these two extremes.
One type of rational but secondary approach is the systematic expansion of known leads. For example, if one antifolic acid compound shows a definite effect, the chemists are asked to synthesize a wide range of chemical relatives or analogues. The principle of metabolic interference is also applied not only to vitamins, but to purines, pyrimidines and other chemicals that are known to be involved in cellular synthesis and metabolism. This approach has yielded some valuable additions, such as 6-mercaptopurine. The antimetabolite area is attractive to the research worker because it contributes to the understanding of cellular metabolic processes even if the aim of anticancer effect is not reached.
Nitrogen mustard is one of several hundred very reactive alkylating chemicals, some effects of which in tissues resemble those of ionizing radiation so that they are referred to as “X-ray mimetic.” Like X-rays, these chemicals not only inhibit some forms of cancer but also can produce cancers, and also cause germinal mutations. The alkylating “war head” can be attached to amino acids, sugars, and other chemicals of metabolic importance, with the hope that these will guide the reaction to a specific target.
Organic chemists have also created hundreds of steroids and related compounds with various spectra of hormonal activity, as potentially useful agents for some forms of cancer.
The choice of materials is not limited to defined chemicals. Several antibiotic products of fungi and bacteria, such as actinomycin and mitomycin, have been found to have cancer inhibiting activity. A systematic exploration of cultures of microorganisms for such effects became a natural part of the program. Crude but defined products of plants have led to the isolation of yet another group of active anticancer agents such as vinblastine and vincristine, from the periwinkle plant (Vinca rosea).
The second phase of the chemotherapy program involves the testing of the selected materials on animals with tumors. Here the investigators are confronted with a number of difficulties. How similar are the cancers of mice or rats and those of man? With what assurance can we predict that an agent that shows an effect on some animal with a tumor will have an effect in a patient with cancer? To answer this question, a large amount of data are necessary for analysis, and probably no single answer will ever be forthcoming. It would be really too much to ask that any one, or a few, types of cancers in animals would have the same responses as the hundred odd different types of cancer in man. A practical approach to this problem was to select the animal tumors that on previous experience paralleled human tumor responses. Despite the known short-comings of transplanted tumors, these are still the ones that are most uniform in their growth and other characteristics, and are obtainable in sufficient numbers for the large scale programs. With additional experience and correlations with clinical trials, less useful animal tumors are replaced in the system. Induced and spontaneous tumors in rodents are also used for agents of particular interest, as well as human cancers grown in cheek pouches of hamsters and in rats that are given cortisone to depress their immune response.
With the analysis of a large body of data, there is now good reason to base chemotherapy investigations directed against human leukemia on several transplanted leukemias in mice, such as the L 1210 tumor. It is inevitable that better experimental models for more limited types of human cancers will be found and will allow more rapid and more confident choices of materials for actual human trial.
The third step in the investigational process is preclinical pharmacology. Agents that affect animal cancers adversely or prolong the life of the animals, or show other beneficial effects, have to be tested further for safety and for determination of routes of administration, dose levels, schedules, and the vehicles for use in man. Acute toxicity is usually determined first in rats, and is extended to acute and chronic toxicity studies in rats and dogs, with other species including monkeys as considered indicated. Very important are the investigations of the metabolic pathways and mode of action of the chemicals. These studies are greatly facilitated by radioactive-isotope tagged chemicals, making it easier to detect the absorption, deposition, excretion and reactions of the agent within the body. Even the most extensive research preparation on animals, however, does not completely eliminate hazard, untoward reactions and other risks when the agent is finally selected for exploration in man. Of course, when we deal with a fatal disease such as advanced cancer, somewhat earlier clinical trial is justified than would be, for example, for the common cold. Nevertheless, in the last analysis the decision to perform a clinical trial rests on the consideration of all available data on the chemistry, action, toxicology and pharmacology by a technically competent group of investigators from the clinical and biological sciences.
Clinical trials in cancer proceed through three phases. The first is the cautious administration to patients of a fraction of a dose of material that has been found to be safe in rats, dogs and other animals. The dose is then gradually increased, and all appropriate observations for toxic or beneficial reactions are recorded and analyzed. At this point the drugs are tried in patients with very advanced cancer who have exhausted the possibilities of all standard and useful medical measures. Their participation in the trials is voluntary and with their full consent. The chief aim is to assure that further trial of the agent is indicated and safe.
Phase two of the clinical trials consists of a more intensive search for anticancer effects in patients with as many types of cancer as possible, unless the agent is designed for, or for other reasons may be predicted as having effects in a limited variety of cancer entities. From a practical standpoint, this phase involves a collaboration among clinical centers, in order to obtain results during a reasonable time. The more common human cancer entities, such as cancers of the lung, large intestine, breast and leukemia, are chosen for the same reason. Again, only patients beyond possible cure or benefit by known procedures are accepted, and the patients are informed and they agree to participate in the studies. The most informative patients are those with measurable, visible lesions, allowing frequent quantitative observations on the course of the disease. This phase establishes whether the agent is ineffective or does have sufficient objective effects on cancer to justify further clinical use.
Phase three of the clinical trials is usually limited to drugs or procedures that have been established as having a beneficial effect. They attempt to quantitate the response, and to compare it with another drug or procedure. A typical question such trials answer is whether nitrogen mustard or a newer alkylating agent, at specific dose schedules, is superior for the management of a particular type of cancer, such as Hodgkin’s disease. In cancers for which no recognized beneficial treatment is available, the control group may receive an inert “placebo.” The best design for these studies is one in which neither the patient nor the physician knows whether the patient is getting one or another drug. This is known as a “double-blind” method. The inclusion of patients is still on a voluntary basis, but the whole purpose of the study would be defeated if the nature of the treatment were known by the individual patient. Thus, patients enter one or another of the groups to be compared by laws of chance, through a procedure called “randomization,” which allows for an equal distribution of known and unknown differences between patients and avoids selection.
As we have emphasized under an earlier section on treatment, there are no curative chemical agents for cancer. Perhaps the closest to this ultimate goal are the long, often complete remissions achieved with antifolics in choriocarcinoma and with castration and female sex hormones in prostatic cancer. Thus, truly effective general drugs for cancer are still in the future. But the definite progress already recorded in this international scientific adventure gives us good grounds for continuing the efforts.