Extended Data Fig. 10: CTP and the long term impact of competing treatments.

While the mean CTP captures a treatment’s short term impact, the complete CTP distribution, determines its long term effect. (a) Two competing hypothetical drugs (blue) and (red) are demonstrated. While the mean CTP of drug B (distribution peak) is smaller, indicating a rapid decline in the cancer cell population within the first week, the tail of this distribution is broader, and hence it exhibits a higher variance in CTP compared to drug A. (b) The calculated cell count, vs. time under the two treatments is presented. At day 7, drug B seems superior, however, as a result its broad tailed distribution, after approximately 2 weeks of treatment it becomes less effective than drug A. Such crossover illustrates the importance of the CTP distribution, as opposed to just the mean, as a predictor of treatment effectiveness. (c) An empirical example showing the cancer cell population under doxorubicin (blue) vs. trametinib (red). While the latter is three times more effective at day 7, the trend is reversed in later time points. (d) Similar crossover is observed for cisplatin vs. trametinib.