The search for new drugs ideally requires an accurate system for prediction of therapeutic effectiveness of candidate compounds in patients, an accurate system for prediction of toxicity, and a knowledge of the relation of chemical structure to both anticancer activity and toxicity. While none of these conditions has been fully met, the National Cancer Institute's comprehensive cancer chemotherapy program, in operation since 1955, has produced enough quantitative data in animals and in patients to define increasingly sophisticated approaches to the identification and characterization of effective agents.

Screening is the process of selecting from the many synthetic and natural chemicals available a few promising compounds with a high probability of therapeutic effectiveness and safety in patients.

The Primary Screen

First step in the search for new drugs is the screening of new materials in a laboratory animal system. About 15000 new materials are screened annually in the National Cancer Institute's comprehensive cancer chemotherapy program: about 15 compounds, or 0.1 percent, show sufficient activity in the primary screen to be retained for further testing as potential cancer drugs.

One of the elements in the screening process is the primary screen, which yearly selects 15 compounds from about 15,000 new materials, for further testing as potential cancer drugs. The primary screen for the past several years has consisted of two laboratory animal systems: L1210 leukemia in the mouse and Walker 256 in the rat. A coded test of 42 compounds known to be clinically active or inactive showed that this screen selected correctly 29 of 30 of the actives and 10 of the 12 inactive compounds. Thus, there were one false negative and two false positive selections.

Research on improving the predictive ability of the primary screen

Research on improving the predictive ability of the primary screen is a continuing process. For example, consideration has been given in the past year to removing Walker 256 temporarily from the screen, since it has not been yielding as broad a spectrum of agents as desirable. Other laboratory animal systems being evaluated include P-388 leukemia, which is sensitive to several materials of potential interest, particularly an antibiotic, mithramycin, and other natural products; and spontaneous AKR leukemia.

Of about 15,000 agents obtained for screening in a year, half may be synthetic chemicals, including hormonal compounds, and the other half, fermentation products and plant and animal products. Natural products represent a source of high potential that has barely been tapped for substances to test in the primary screen. As a consequence, interest is focusing on obtaining extracts from algae and other marine plants, and from animals, insects, and higher plants from different geographical areas, as well as pure materials already isolated from natural products by investigators throughout the world.

Cancer drug investigators are interested in natural products as a source of high potential that has barely been tapped for substances to test in the primary screen. Extracts are being obtained from algae and other marine plants, and from animals, insects, and higher plants from different geographical areas.

While random sampling of materials for testing in the primary screen continues, efforts to relate chemical structure to anticancer activity are in progress. Structure-function relationships among folic acid antagonists and nucleosides have thus far been studied. It is hoped that in the future the choice of more and more compounds introduced into the screen will be determined by structure-function considerations. These compounds would include some in which the molecule of natural products was altered, as well as new synthetic structures designed for producing anticancer activity.

Toxicity

Another element in the screening process concerns the prediction, on the basis of animal studies, of the quality and severity of toxic effects that can be expected to attend clinical use of a new drug. Because of precise data available from animal studies and increasingly better understanding of the relation of such data to the clinical situation, prediction of starting doses of new drugs has been made more reliable than formerly. Qualitative prediction of drug toxicity refers to the specific organ damage to be expected when drugs are given to patients, and except for skin and nervous system effects, generally may be accurately inferred from animal data. Research is in progress to find an improved predicting system through which nervous system toxicity can be accurately predicted for man.

Laboratory research of the past several years on cell kinetics has yielded significant information on selective toxicity for the cancer cell. Efforts are pressing forward to classify drugs with regard to their effects on different phases of the cell cycle and to determine their specific toxicities for normal and tumor cells (page 62).

Expanded pharmacologic studies of active drugs, comparing the disposition of the drugs in animals and man, have yielded important new knowledge in the past year. Cyclophosphamide, one of the drugs studied, is an excellent cancer drug in animal systems but, although a very good drug in clinical cancer, its performance falls short of the animal model (page 58). Studies showed that cyclophosphamide itself is essentially inactive as an antitumor agent. When it is taken into the body, the resulting metabolites are the active antitumor compounds. They are present in high concentrations in the mouse and rat, but in low concentrations in patients. Furthermore, the possibility exists that unaltered cyclophosphamide may play a role in its toxicity. Other studies showed that when the metabolism of the compound was blocked in mice, the toxicity was increased and the mice died more rapidly than control mice. Further studies of the active metabolites are in progress.

cyclophosphamide tests

Studies of the fate of drugs in the body are an important part of the search for new cancer agents. This illustration suggests why cyclophosphamide, an active drug, has much more activity in the laboratory mouse and rat than in man. Studies showed that cyclophosphamide itself is essentially inactive. When it is taken into the body, the resulting metabolites are the active antitumor compounds and the concentrations of these metabolites are high in the animals but low in patients.