Skip to main content
. 2017 Mar 13;7:44206. doi: 10.1038/srep44206

Figure 5. Exploring the robustness of treatment strategies through model simulation.

Figure 5

(A) Switching strategies are more beneficial to tumor cell populations with more initial heterogeneity. (Left) Fold change in final lung adenocarcinoma tumor cell populations at day 30 versus day 0 over the course of the optimal 30, 15, 10, 5, 3, and 1 day treatment strategies solved by algorithm 1 (SI, Section 2.2) and normalized by fold change in final tumor cell population for the constant 30 day treatment strategy for an initial tumor cell population comprised of (90% EGFRL858R, 10% H1975 EGFRL858R, T790M) and another comprised of (89% EGFRL858R, 10% BRAFV600E, 1% EGFRL858R, T790M) subclones. (Right) Sum of fold change for the final lung adenocarcinoma populations (SI, Equation S5) for select initial tumor cell distributions (SI, Table 1) and their corresponding optimal 30, 15, 10, 5, 3, and 1 day treatment strategies, categorized by the number of subclones in the initial tumor cell population. Smaller fold change sums indicate that more switching is beneficial to reduce final populations, whereas larger fold changes indicate that more switching does not necessarily help in reducing the final tumor populations. (B) EGFR TKI dose perturbations. (Left) Fold change in number of lung adenocarcinoma cells between day 30 and day 0, as a function of percent EGFR TKI dose reduction for the optimal 30, 15, 10, 5 and 1 day strategies solved by algorithm 1 (SI, Section 2.2) for tumor cell populations indicated above. The shaded areas indicate the regions of the perturbation space where the treatment strategy reduces the initial tumor cell population by more than 30% (response, light blue), increases the size of the original tumor population size by more than 20% (progression, red), or maintains the original tumor population size between the two (stability, white). (Right) Bar graphs indicate the maximum reduction in EGFR TKI dose supported by the optimal strategy such that there is still reduction in tumor size at day 30 with respect to day 0 for the V600E and the pretreatment MET tumor. (C) The average maximum percent EGFR TKI dose reduction supported before progression for lung adenocarcinoma tumors with different number of initial tumor cell subpopulations and for predicted optimal 30, 15, 10, 5, and 1 day switching strategies.