A simulation of cell kinetics with application to L-1210 cells

The normal cell cycle is described and a model developed which postulates that the length of stay in each state is a stochastic variable following the normal or gamma distribution. This model permits the evaluation of different schedules of treatment on the proliferation of cells; each treatment is assumed to have a specified probability of killing a cell in the various states and this probability remains applicable for the period of activity of the drug. The proliferation of L-1210 cells in mice has been simulated on the computer and the mean and variability of time to death obtained corresponds very closely with the experimental results of Skipper, Schabel, and Wilcox (Cancer Chemotherapy Reports, 35:1964) if gamma distributions are assumed for the lengths of stay in each state and if the logarithm of the number of cells required to kill an animal is assumed to follow an exponential distribution with a median of 1.5 x 10[superscript 9]. The effect of different schedules of treatment with cytosine arabinoside on the proliferation of L-1210 cells was investigated by simulation on the computer. Median life spans of mice receiving one treatment of cytosine arabinoside could be matched but results of multiple doses could not be simulated under the same set of assumptions. The increase of life span after several doses of the same amount of drug did not show a simple relationship to the increase in life span caused by one dose. This suggests further experimentation to measure the lengths of stay in the various states of the cell cycle after multiple administrations of drug.