Mathematical modeling of hypoxia and adenosine to explore tumor escape mechanisms in DC-based immunotherapy

Abstract

Identifying and controlling tumor escape mechanisms is crucial for improving cancer treatment effectiveness. Experimental studies reveal tumor hypoxia and adenosine as significant contributors to such mechanisms. Hypoxia exacerbates adenosine levels in the tumor microenvironment. Combining inhibition of these factors with dendritic cell (DC)-based immunotherapy promises improved clinical outcomes. However, challenges include understanding dynamics, optimal vaccine dosages, and timing. Mathematical models, including agent-based, diffusion, and ordinary differential equations, address these challenges. Here, we employ these models for the first time to elucidate how hypoxia and adenosine facilitate tumor escape in DC-based immunotherapy. After parameter estimation using experimental data, we optimize vaccination protocols to minimize tumor growth. Sensitivity analysis highlights adenosine's significant impact on immunotherapy efficacy. Its suppressive role impedes treatment success, but inhibiting adenosine could enhance therapy, as suggested by the model. Our findings shed light on hypoxia and adenosine-mediated tumor escape mechanisms, informing future treatment strategies. Additionally, identifiability analysis confirms accurate parameter determination using experimental data.

Keywords: Adenosine; Dendritic cell-based immunotherapy; Hypoxia; Mathematical modeling; Tumor escape mechanisms.

Figure 1

The suggested model’s overall structure. In this model, tumor, DC, T cell, Treg, effector cell, and factors such as hypoxia and adenosine are considered. The interaction of the factors with each other is shown by the arrows and their corresponding parameter is specified.

Figure 2

Interactions and communication between tumor and immune system cells in the discrete grid over time period Δt. The tumor cell activates its neighboring inactive DC (A1–B1). The T cell next to the active DC becomes a Treg (A3–B3) or effector cell (A5–B5) after activation. The effector cell kills the tumor cell adjacent to it (D1–D2). The Treg cell inactivates its neighboring active DC (D5–E4).

Figure 3

Tumor volume was predicted by the model and measured in the empirical data^, in different groups. (a) Untreated. (b) Treatment with adenosine inhibitor. (c) Treatment with DC vaccine. (d) Treatment with adenosine inhibitor and DC vaccine. (e) Treatment with hypoxia inhibitor. (f) Treatment with hypoxia inhibitor and DC vaccine. (g) Treatment with adenosine and hypoxia inhibitor. (h) Treatment with adenosine and hypoxia inhibitor and DC vaccine.

Figure 4

The number of effector cells and Treg cells in different groups. (a) Untreated. (b) Treatment with adenosine inhibitor. (c) Treatment with hypoxia inhibitor. (d) Treatment with adenosine inhibitor and hypoxia inhibitor.

Figure 5

Comparison of the oxygen concentration and tumor cells in the untreated and siRNA + Px-478 + DC groups on day 12. (a) Oxygen concentration in the untreated group on day 12. (b) Oxygen concentration in the siRNA + Px-478 + DC group on day 12. (c) Tumor cells in the untreated group on day 12. (d) Tumor cells in the siRNA + Px-478 + DC group on day 12.

Figure 6

Investigating the time course of effector cell cytotoxicity rate and adenosine concentration in four groups. (a) Effector cell cytotoxicity rate, (b) the adenosine concentration.

Figure 7

The average number of effector cells around the tumor as a result of changing the dose of the DC vaccine.

Figure 8

Sensitivity value of the model parameters.

Figure 9

The time course of tumor volume, effector cell cytotoxicity rate, adenosine concentration, and number of Treg as a result of changes in β (a) Tumor volume. (b) Effector cell cytotoxicity rate. (c) Level of adenosine. (d) Number of Treg.

Figure 10

The time course of tumor volume as the parameters µ and P_Treg change. (a) Tumor volume as a result of the change in µ. (b) Tumor volume as a result of the change in P_Treg.