One of the results of studies of cell populations was the finding that a given dose of a drug killed the same percentage of tumor cells, regardless of the size of the cell population. For example, the same dose of a drug kills 99 percent of 100,000 leukemic cells or 99 percent of 1,000,000 leukemic cells, whichever number is present in the host individual. While the number of cancer cells destroyed is larger if the population is 1 million cells than if it is 100,000 cells, the fraction destroyed remains the same.
Research has shown that a given dose of a drug kills the same percentage of tumor cells, regardless of the size of the cell population. In leukemic mice, cures can be produced if the drug is administered on an intermittent schedule to allow for recovery of normal cells and if the dose is large enough so that the percentage of the cell population killed outpaces the multiplication of surviving leukemia cells.
For effective treatment the percentage of the cell population killed by the drug must be high enough so that multiplication of surviving cancer cells will not outpace the inhibiting effect of the drug and kill the host. In theory, a number of periods of treatment and recovery would consecutively reduce the number of cancer cells and they would be essentially eliminated. Obviously, drug treatment is most effective when given early in the course of a disease or as an adjunct to surgery, when tumor cells are comparatively few in number. The procedure is based on the assumptions that the fractional reduction of tumor cells will be constant each time the drug is given, that host recovery will be constant between treatments, and that the drug will reach all cells in the body.
Another finding was that, with single doses of drugs, the higher the dosage the greater the percentage kill of tumor cells.
A third finding related to the importance of spacing the administration of a drug for achieving a high level of tumor cell kill. The lifespan of leukemic mice was increased when they were treated with methotrexate once every fourth day rather than daily. This principle was applied clinically and found to be similarly effective. In a group of children with acute lymphocytic leukemia brought into remission with a combination of prednisone and vincristine sulfate, survival was significantly longer for those receiving maintenance doses of methotrexate intramuscularly twice a week than for the ones who received the drug orally each day. The investigators conducting this study reported that the median duration of remission was 17 months for the former group and 3.3 months for the latter.
From a series of continuing laboratory experiments it has become apparent that increasing doses of some drugs need not kill increasing fractions of normal cells but do destroy increasing fractions of tumor cells. The evidence has suggested that a certain percentage of normal bone marrow stem cells (most primitive cells in the bone marrow) is in a nondividing stage when some drugs do not affect them. Tumor cells seem not to have such a phase and are therefore more susceptible to the toxic effects of the drugs. Thus, transplanted lymphoma in mice and rapidly proliferating marrow cells were six times as sensitive to 5-fluorouracil as were normal colony-forming stem cells in the marrow, which were not proliferating.
Laboratory studies have shown that some drugs do not attack cells when they are in a nondividing stage. Thus, in mice, transplanted lymphoma cells, which were rapidly proliferating, were several times more sensitive to 5-fluorouracil than were normal colony-forming stem cells in the bone marrow, which were not proliferating.
In another experiment, dactinomycin and cyclophosphamide were found to be similar to 5-fluorouracil in their action under the conditions of the study. These drugs had a six- to ten-fold greater effect on the lymphoma cells than on the normal cells and killed cells throughout all phases of their generation cycle. Sensitivity of the cell populations, it was concluded, depended on the fraction of cells in the proliferative state.
Vinblastine and methotrexate had a different mode of action in these experiments. At high doses, a large fraction (20 to 60 percent) of the normal hematopoietic colony-forming cells survived, whereas only a very small fraction (0.05 to 0.08 percent) of the lymphoma cells survived. This marked difference in response was attributed to the probability that these agents kill only those cells in one portion of the generation cycle. Lymphoma cells complete a generation cycle about once every 10 hours, so that vinblastine, which kills cells at or near mitosis, would destroy the proliferative capacity of most tumor cells; similarly methotrexate which kills cells in the DNA synthesis (S) phase would also destroy most lymphoma cells. On the other hand, the normal cells pass through the generation cycle infrequently and therefore only a small fraction would be likely to be proliferating and sensitive to the drugs in the 24-hour period of treatment used in the study.
In a third group in this series of experiments were nitrogen mustard and gamma-irradiation, which killed normal and tumor cells with the same efficiency. The action of these agents apparently does not depend strongly on the proliferative state of the exposed cells.