Processes 2019, 7,37
some studies on diseases other than cancer where we feel that the methodology is relevant and could
be applied to cancer; we also refer to a few general software tools that can be readily adapted to cancer.
We break the review into the following sections, although there may be significant overlap:
(1) Models focusing on immune-related tumor mechanobiology
(2) Models focusing on tumor-associated vasculature in the immune response
(3) Models focusing on tumor-associated lymphatics and lymph nodes
(4) Models focusing on tumor immunotherapy
(5) Models focusing on tumor-enhancing immune cells
(6) Models focusing on intra-tumor heterogeneity
3.1. Models Focusing on Tumor Mechanobiology
Changes in the tumor extracellular matrix (ECM) have been known to contribute to tumor
progression and metastasis [85], with several computational models focusing on investigating glioma
invasion [86–88], but less is known about its contribution to immune response. Computational
modeling has been used to shed light on the interactions between the ECM and the immune system in
cancer dynamics. A hybrid agent-based model was used to investigate the role of cellular adhesion
to the ECM in tumor and immune system dynamics [89]. Frascoli et al. found that the greater the
motility of the cancer cells, the more likely they will escape from immunotherapy. They also found
that intermediate levels of adhesion in general led to less successful outcomes, but these results
were variable.
Kather et al. used ABM to investigate the combination of adoptive cell transfer and therapy that
permeabilized the fibrotic stromal component in colorectal cancer [72]. Adoptive cell transfer is a
therapeutic strategy that aims to increase the number of immune cells to strengthen immuno-surveillance
and counter tumor development. Kather et al. simulated various conditions of immune surveillance. In
their model, T-cell killing of tumor cells occurred in a purely stochastic manner, with killing probability
representing the effect of tumor specificity, immunogenicity, stimulatory and inhibitory effects all in
one parameter. An immune rich environment promoted immune escape, but tumor growth slowed
in a lymphocyte deprived environment. Tumor control was observed in a subgroup of tumors with
less stroma and a high numbers of immune cells. They found that high levels of fibrosis and low
numbers of lymphocytes reduced overall survival. Their findings were validated with data from
colorectal cancer patients, where low density stroma and high lymphocyte level correlated with better
overall survival. In this study, Kather et al. simulated the effect of immunotherapy by boosting the
number of immune cells by 2–8 fold. Therapy was intended to enhance fibrotic stromal permeability;
this was implemented by modifying the corresponding parameter by a factor of 4% to 16%. The
model predicted that optimal tumor eradication requires a combination of therapeutics aiming at both
activating adaptive immune system and stromal depletion.
3.2. Models Focusing on Tumor-Associated Vasculature in the Immune Response
Tumor-associated vasculature is an important aspect of the tumor-immune complex because
it not only provides oxygen and nutrients for the tumor to grow, but it is also the source of tumor
dissemination via circulating tumor cells (CTC), and recruitment for many immune cells, such as
monocytes/macrophages and T-cells. Studies have aimed to provide a better understanding of
these processes [90,91]. An ABM of Early Metastasis (ABMEM) framework was used to model the
interactions between tumor cells, platelets, neutrophils, and endothelial cells [92]. Receptor binding to
Mac-1 (macrophage antigen-1) by endothelial cells, platelets, or tumor cells leads to reactive oxygen
species (ROS) production by neutrophils. Uppal et al. examined two types of platelet inhibition:
inhibition of thromboxane, inhibition of adenosine diphosphate (ADP) receptors and inhibition of
both [92]. They found that thromboxane inhibition alone resulted in the best outcome.
46
some studies on diseases other than cancer where we feel that the methodology is relevant and could
be applied to cancer; we also refer to a few general software tools that can be readily adapted to cancer.
We break the review into the following sections, although there may be significant overlap:
(1) Models focusing on immune-related tumor mechanobiology
(2) Models focusing on tumor-associated vasculature in the immune response
(3) Models focusing on tumor-associated lymphatics and lymph nodes
(4) Models focusing on tumor immunotherapy
(5) Models focusing on tumor-enhancing immune cells
(6) Models focusing on intra-tumor heterogeneity
3.1. Models Focusing on Tumor Mechanobiology
Changes in the tumor extracellular matrix (ECM) have been known to contribute to tumor
progression and metastasis [85], with several computational models focusing on investigating glioma
invasion [86–88], but less is known about its contribution to immune response. Computational
modeling has been used to shed light on the interactions between the ECM and the immune system in
cancer dynamics. A hybrid agent-based model was used to investigate the role of cellular adhesion
to the ECM in tumor and immune system dynamics [89]. Frascoli et al. found that the greater the
motility of the cancer cells, the more likely they will escape from immunotherapy. They also found
that intermediate levels of adhesion in general led to less successful outcomes, but these results
were variable.
Kather et al. used ABM to investigate the combination of adoptive cell transfer and therapy that
permeabilized the fibrotic stromal component in colorectal cancer [72]. Adoptive cell transfer is a
therapeutic strategy that aims to increase the number of immune cells to strengthen immuno-surveillance
and counter tumor development. Kather et al. simulated various conditions of immune surveillance. In
their model, T-cell killing of tumor cells occurred in a purely stochastic manner, with killing probability
representing the effect of tumor specificity, immunogenicity, stimulatory and inhibitory effects all in
one parameter. An immune rich environment promoted immune escape, but tumor growth slowed
in a lymphocyte deprived environment. Tumor control was observed in a subgroup of tumors with
less stroma and a high numbers of immune cells. They found that high levels of fibrosis and low
numbers of lymphocytes reduced overall survival. Their findings were validated with data from
colorectal cancer patients, where low density stroma and high lymphocyte level correlated with better
overall survival. In this study, Kather et al. simulated the effect of immunotherapy by boosting the
number of immune cells by 2–8 fold. Therapy was intended to enhance fibrotic stromal permeability;
this was implemented by modifying the corresponding parameter by a factor of 4% to 16%. The
model predicted that optimal tumor eradication requires a combination of therapeutics aiming at both
activating adaptive immune system and stromal depletion.
3.2. Models Focusing on Tumor-Associated Vasculature in the Immune Response
Tumor-associated vasculature is an important aspect of the tumor-immune complex because
it not only provides oxygen and nutrients for the tumor to grow, but it is also the source of tumor
dissemination via circulating tumor cells (CTC), and recruitment for many immune cells, such as
monocytes/macrophages and T-cells. Studies have aimed to provide a better understanding of
these processes [90,91]. An ABM of Early Metastasis (ABMEM) framework was used to model the
interactions between tumor cells, platelets, neutrophils, and endothelial cells [92]. Receptor binding to
Mac-1 (macrophage antigen-1) by endothelial cells, platelets, or tumor cells leads to reactive oxygen
species (ROS) production by neutrophils. Uppal et al. examined two types of platelet inhibition:
inhibition of thromboxane, inhibition of adenosine diphosphate (ADP) receptors and inhibition of
both [92]. They found that thromboxane inhibition alone resulted in the best outcome.
46