# Indirect Contributions to Tumor Dynamics in the First Stage of the Avascular Phase

## Abstract

**:**

## 1. Introduction

## 2. Theory and Method

^{3}a fixed volume occupied by the tissue, having S as its bounding surface and n as the number of interacting species in Ω; at each time step t ∈ [0,T] and position

**x**= (x,y,z) ∈ Ω, c

_{i}= c

_{i}(

**x**,t), for i = 1, …, n, represents the concentration of the ith species in Ω. Then, introducing c = (c

_{1}, …, c

_{n}), the mass conservation law for each species gives:

_{i}(

**c**) is a source term taking into account contributions coming from all the interacting species in Ω, and

**φ**

_{i}(

**x**,t) is the flux of c

_{i}through S. From Equation (1), using the divergence theorem, the evolution equation for the ith species is obtained:

_{i}(

**x**,t) corresponds, respectively, to cancer cells (c), VN (v), uPA (u), PAI-1 (p), and plasmin (m).

_{c}; haptotaxis due to VN and chemotaxis due to uPA and PAI-1, are accounted for by the χ

_{v}, χ

_{u}and χ

_{p}coefficients, respectively; cell production is accounted for by the logistic term through the μ

_{1}coefficient. We next turn our attention to PAI-1/VN interaction near the cancer cell’s surface, since it is relevant for the mass conservation of cancer cells. In Chaplain and Lolas [21] it was observed that such a bond indirectly promotes cancer cell proliferation. The mechanism starts as a signaling triggered by PAI-1/VN binding; PAI-1 molecules removed by VN result in the activation of the uPA to cancer cell binding through the uPA receptor (uPAR) located at the cell surface, which in turn activates plasmin formation, which can be considered a PAI-1/VN by-product fostering cancer proliferation. Then, we account for such indirect proliferation mechanism with the ϕ

_{14}coefficient:

_{0}is intended as the carrying capacity of the cancer cell population, and will be defined later on, also for scaling purposes. The ECM representative VN macromolecule is confined within the cellular environment, and then the relative flux term vanishes giving the model equation:

_{21}: such binding, in fact, facilitates the VN binding to uPAR, then initiating a signaling promoting its own production [21,22,52]. A logistic proliferation is also present at rate μ

_{2}, while VN consumption, instead, is due to its neutralization by PAI-1 and its degradation by plasmin, at rates ϕ

_{22}and δ, respectively. Similarly to cancer cell population, in Equation (4) the constant v

_{0}is intended as the carrying capacity of the vitronectin population, which will be used for scaling purposes later on.

_{u}, and is produced by cancer cells at a rate α

_{31}; moreover, it is inhibited via interaction with PAI-1 and degraded by interaction with cancer cells, at rates ϕ

_{31}and ϕ

_{33}, respectively:

_{p}, a term accounting for PAI-1 production by healthy cells at a rate α

_{41}is present, triggered by chemical signaling starting from plasminogen activation [21]. PAI-1 degradation comes from its interaction with uPA and VN at rates ϕ

_{41}and ϕ

_{42}, respectively:

_{m}, it is indirectly promoted by the PAI-1/VN and cancer cells/uPA interactions at rates ϕ

_{52}and ϕ

_{53}, respectively. In fact, uPA binds to its uPAR receptor, which is located at the cell surface, where proteolytic activity increases due to uPA plasminogen activation to plasmin [21,22,52]. On the other hand, as in the case of the cancer cells equation, the binding PAI-1/VN promotes the interaction of uPA with cancer cells via the uPAR, inducing indirect plasmin formation through the activation of plasminogen to plasmin [21,22]. Plasmin degrades globally [22] at a rate ϕ

_{54}. On the other hand, the PAI-1/uPA binding prevents the activation of plasminogen to plasmin by uPA, resulting in indirect inhibition of plasmin formation [21]; we then introduce such a contribution through the ϕ

_{51}coefficient:

**x**,0) = exp(−|

**x**|

^{2}ε

^{−1}), v(

**x**,0) = 1 − 0.5c(

**x**,0), u(

**x**,0) = 0.5c(

**x**,0), p(

**x**,0) = 0.05c(

**x**,0), and m(

**x**,0) = 0, where ε = 0.01 mm

^{2}. In other words, we define that at t = 0 the domain occupied by the biological tissue is almost filled by the ECM, and an initial cluster of cancer cells already exists. Chemical signaling in the biological environment starts the production of uPA and PAI-1 proportionally to cancer cells, while plasmin will be produced later, during the evolution process. Since we simulated the cancer cell proliferation in the early stage of the tumor formation, we posited that all the system components stay confined within the biological domain, and then we applied zero-flux boundary conditions to the domain boundaries. The spatial FEM discretization was performed using the method of lines, with a uniform, mapped mesh with 10

^{4}elements, over a two-dimensional spatial domain with a 1 mm

^{2}area, while for time discretization we used an implicit Euler method. Moreover, the local variable distribution was interpolated with quadratic shape functions, using Galerkin’s method for the residuals of differential equations [58].

_{0}= 6.7 × 10

^{7}cell cm

^{−3}, and v

_{0}= 1 nM, respectively; for uPA, PAI-1 and plasmin, instead, u

_{0}= 1 nM, p

_{0}= 1 nM, and m

_{0}= 0.1 nM were used, respectively. We also obtained from the literature all of the parameters used in the model [21,22], markedly the reference value D = 10

^{−6}cm

^{2}s

^{−1}used to scale the diffusion coefficients and the reference distance for cancer cells in the early stages of invasion, L = 0.1 cm, used to scale the spatial dimensions. It follows that defining τ = L

^{2}D

^{−1}= 10

^{4}, time is scaled accordingly. Cancer cell proliferation indirectly induced by PAI-1/VN interaction was monitored imposing ϕ

_{14}∈ [0, 0.275, 0.55]. This choice was determined after noticing that the parameter values for the PAI-1/VN interaction fall within the non-dimensional range 0 ÷ 0.55 [21,22,52,53]. Moreover, in [21], the authors used a plasmin deactivation rate indirectly induced by the PAI-1/uPA interaction falling within the non-dimensional range 0.15 ÷ 0.75; we thus chose the parameter ϕ

_{51}∈[0, 0.375, 0.75]. The other parameters used in the model simulation are summarized in Table 1, first column, while in the third and fourth columns their non-dimensional forms and non-dimensional values are reported, respectively. We performed our computations over a two-dimensional biological tissue, discretizing the domain for a x ∈ [0,1], y ∈ [0,1] non-dimensional interval, for t ∈ [0,100] corresponding to about 11.6 days, with a time step size δt = 0.1.

## 3. Results and Discussion

_{51}= 0, i.e., considering that plasmin deactivation indirectly induced by the PAI-1/uPA interaction is initially zero, and for ϕ

_{14}∈ [0, 0.275, 0.55]. In Figure 2, Figure 3 and Figure 4 the concentration maps over the simulated domain are shown for, respectively, cancer cells, VN, and plasmin, with snapshots taken at t = 40 (about 4.6 days), 60 (about 6.9 days), 80 (about 9.3 days), and 100 (about 11.6 days), as labeled at the beginning of each panel row; in each column, instead, panels are grouped according to the labeled ϕ

_{14}value above them. Concentrations are linearly mapped on a color scale between the blue and yellow colors. At t = 0 the variables’ distribution is according to the initial conditions, hence they are omitted. The above graphical scheme will be maintained throughout the paper.

**x**= (0,0) grew, and the invasion front developed while propagating with a sharp wave-front inside the squared domain, with a slight increasing invasion speed from ϕ

_{14}= 0 to ϕ

_{14}= 0.55; at t = 60 the invasion front reached the domain boundaries crushing on them, and at the same time a faintly visible secondary invasion front started propagating near

**x**= (0,0). Because of the boundary conditions imposed, all of the interacting species remained confined into the domain; hence from t = 80 to t = 100, while the first invasion front stayed crushed on the domain boundaries, the secondary invasion front started degrading itself in a diffuse and heterogeneous pattern of cancer cells. At t = 80 we observe cancer cells heterogeneously arranged in a network-like structure clustering with growing ϕ

_{14}values. At t = 100 the cancer cell distribution appeared branched with ϕ

_{14}= 0, evolving towards denser and more heterogeneous cell clusters for increasing ϕ

_{14}values. Cancer cells are spread out over the whole domain, even if their nucleation is not clearly visible on the boundary surfaces. Additionally, a symmetry of the heterogeneities spatial distribution with respect to the diagonal, starting from the domain origin, is observed, due to the initial conditions imposed for the solution of the PDE system.

_{14}parameter. In fact, the VN distribution was eroded at the domain boundaries and near

**x**= (0,0), while inside the simulated domain the VN diffusion increased according to the ϕ

_{14}value. At t = 80, nevertheless, the VN maps unexpectedly showed a growing concentration according to the cancer cell clustering (see Figure 2) where instead VN should have been consumed. Such a behavior in part persisted at t = 100, though it was limited to certain regions of the simulated domain, particularly for ϕ

_{14}= 0. This matter turns clearer looking at the plasmin concentration of Figure 4: in each panel, the map overlaps with the corresponding map of Figure 2, showing diffuse patterns related to the activity area of cancer cells, especially at the boundary regions, which are not well visible in Figure 2.

_{51}= 0.375 are shown in Figure 5, Figure 6 and Figure 7 for, respectively, cancer cells, VN, and plasmin. Comparing the concentration maps for cancer cells at t = 40 and t = 60 of Figure 2 and Figure 5, indirect plasmin deactivation did not seem to influence the speed of the invasion front, depending only on ϕ

_{14}. At the higher time steps, instead, the network branching became sharper with respect to Figure 2, with the mesh size reducing faster as ϕ

_{14}increased. Moreover, the cancer cell clusters appeared more densely distributed and interconnected across the domain, both at t = 80 and at t = 100.

_{51}= 0. In addition, the plasmin distribution, shown in Figure 7, which is superimposable to the cancer cell maps of Figure 5, exhibited a diffuse pattern, further evidencing the tumor activity regions inside the simulated domain, especially close to the domain boundaries.

_{51}= 0.75, the distribution maps for cancer cells, shown in Figure 8, were similar to the previous cases, as far as the lower time steps are concerned. For t = 80, instead, the network branching appeared to be slightly affected by the ϕ

_{14}coefficient, while at t = 100 the branching seemed inhibited, particularly at lower ϕ

_{14}values. Concerning VN, as shown in Figure 9, the concentration maps appeared even more definite, while the plasmin concentration maps, shown in Figure 10, always enhanced cancer cell activity regions.

_{51}= 0 to ϕ

_{51}= 0.375 and ϕ

_{51}= 0.75. As cancer cell proliferation occurs at the expense of VN, the suppression of the plasmin deactivation contribution by imposing ϕ

_{51}= 0 should produce a decrease in VN; however, an unexpected and unreasonable growth of VN concentration was observed—see Figure 3 at t = 80 and t = 100. We can thus deduce that contributions coming from indirect plasmin deactivation that is different from zero improves the model and makes it more realistic. During the early stage of the tumor proliferation at the expense of VN, the primary invasion front intensity grew until t = 60, when it crushed on the domain boundaries, and the secondary front started to grow from the original seeding site. Whatever the values of ϕ

_{14}and ϕ

_{51}were, such an occurrence was evidenced in the maps of both VN and plasmin. At this stage, the VN degradation by the primary invasion front was not complete, as witnessed by the presence of the background at t = 60; therefore some residual healthy tissue remained in the domain, on which the secondary invasion front could proliferate. In fact, for t > 60, the domain became completely invaded, and clusters of cancer cells spread and branched all over it, starting to build a network as the secondary invasion front evolved. Nevertheless, the VN maps for t > 60 evidenced that residual healthy ECM was still present, except for domain regions where the degrading activity of cancer cells left a kind of imprint in correspondence to high concentration areas of cancer cells. A similar behavior was observed in [52], where the authors simulated tumor progression in a two-dimensional domain, also exploring the micro-dynamics of cancer invasion; their qualitative drawing of the tumor invasive boundary appeared, on average, to be in agreement with ECM consumption maps, except at higher time steps. By comparison, our results, obtained on a spatial domain four times smaller and using a different model, with different model parameters and a different numerical method, while using the FEM and interpolating the variable distribution inside each element with quadratic shape functions, gave us a numerical result similar to that obtained with the mixed method used in [52] at a lower computational cost.

_{14}coefficient a slight increase of the invasion speed occurs, regardless of the ϕ

_{51}coefficient, meaning that the tumor invasion speed can be regulated by the PAI-1/VN interactions, as observable in Figure 2 and Figure 5, and, at t = 40. Secondly, suppressing the contribution coming from indirect plasmin deactivation does not seem to give reliable results, and therefore leads us to infer that such a contribution needs to be considered. Nevertheless, focusing on the results obtained for ϕ

_{51}= 0.375 and ϕ

_{51}= 0.75, in the former case the invasion patterns showed stronger heterogeneities with respect to ϕ

_{51}= 0.75, at t = 80 and t = 100, as if high plasmin deactivation rates had a limited influence on cancer cell dynamics. Heterogeneity and fast invasion of biological tissues are hallmarks of malignancy [3], which is then supported by increasing rates of PAI-1/VN binding. The PAI-1/uPA binding, instead, seems to have a slower influence on cancer cell dynamics, being more detectable for t = 100: it reverberates on plasmin deactivation, and for ϕ

_{51}= 0.75 results in more stable (t = 80) and less heterogeneous (t = 10) proliferation patterns.

_{53}parameter, induced a lowering of the speed of the invasion front. Our simulations, instead, showed that indirect plasmin deactivation induced by the PAI-1/uPA interaction, therefore decreasing the amount of available plasmin, had no effect on the speed of the invasive front.

_{14}rates increased the speed and heterogeneities of the invasive front; on the other hand, growing ϕ

_{51}rates induced more plasmin deactivation, thus damping the ϕ

_{14}action. Such findings confirm our deductions that the plasmin patterns give an early prediction for cancer activity, and plasmin is itself a marker for tumor formation. In paper [60] it was experimentally shown that PAI-1 overexpression reduces cancer cell migration in vitro, and metastasis in vivo, through uPA inhibition, while the role of the PAI-1/VN binding in tumor invasion appears unclear; moreover, a simultaneous interaction of PAI-1 with both vitronectin and uPA is needed to inhibit metastasis, while the inhibition of tumor growth is primarily due to the uPA inhibitory activity of PAI-1. Our numerical results then help distinguish the separate effects of the PAI-1/VN and PAI-1/uPA bindings, assessing the role of the PAI-1/VN binding in heterogeneous cancer proliferation; further, they answer the question of predicting the spatial extent of tumor formation by observing the plasmin evolution patterns, indicating plasmin as a marker to monitor the heterogeneity of cancer evolution in vivo. Finally, our results suggest a possible method for cancer treatments, designing therapies aimed at supporting plasmin deactivation, and, above all, targeting PAI-1/VN binding. Furthermore, monitoring the plasmin dynamics can give a practical tool to foresee malignant and heterogeneities evolution. In this vein, a new series of simulations are planned using a different geometry in 2D and in 3D, in order to include in the model external contributions as well as effects due to nutrient feeding on cancer evolution.

## 4. Conclusions

_{14}parameter. At this stage, when increasing the ϕ

_{14}parameter from 0 to 0.55, an average speed increase of about 12% was observable in the cancer cell proliferation maps, and this behavior seems independent from the ϕ

_{51}parameter. Once the domain has been invaded, in the presence of a residual ECM, as a consequence of a secondary invasive front cancer cells start to cluster inside the domain and close to its boundaries. From here heterogeneous and symmetric structures start to branch out in a network whose mesh size depends on the ϕ

_{14}parameter. In particular, when increasing the ϕ

_{14}parameter value, the mesh tends to thicken, except for ϕ

_{51}= 0.75 when changes in the ϕ

_{14}value slightly affect cancer heterogeneities. In this respect, it is noteworthy to highlight that at higher ϕ

_{51}values the secondary invasion front activity appeared slightly inhibited. The VN concentration maps can be viewed as a photograph giving a qualitative image of the boundaries of the tumor invasion front. Nevertheless, some inconsistencies occurring in the maps at ϕ

_{51}= 0, at higher time steps, persuade us to consider this value for the parameter as unreasonable. On the other hand, viewing the plasmin as a kind of marker for cancer formation, its concentration maps are revealed to be more reliable in comparison to VN maps for monitoring the cancer activity area. Finally, the computed plasmin concentration maps add useful information about the spatial extension of the tumor activity, highlighting features not clearly visible in the cancer cell concentration maps, delimiting a kind of “confidence interval” for malignant activity. Hence, therapies aimed at targeting the PAI-1/VN binding and at the same time supporting the PAI-1/uPA interaction could be a strategy for effective cancer treatments. In addition, plasmin concentration can be considered as a useful marker for an effective delimitation of the tumor invasive boundary.

## Funding

## Conflicts of Interest

## References

- Preziosi, L. Modelling tumour growth and progression. In Progress in Industrial Mathematics at ECMI 2002; Buikis, A., Ciegis, R., Fitt, A.D., Eds.; Springer: Berlin-Heidelberg, Germany, 2004; pp. 53–66. [Google Scholar]
- Qian, B.Z.; Pollard, J.W. Macrophage diversity enhances tumor progression and metastasis. Cell
**2010**, 141, 39–51. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Hanahan, D.; Weinberg, R.A. The hallmarks of cancer: The next generation. Cell
**2011**, 144, 646–674. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Macklin, P.; McDougall, S.; Anderson, A.R.A.; Chaplain, M.A.J.; Cristini, V.; Lowengrub, J. Multiscale modelling and nonlinear simulation of vascular tumour growth. J. Math. Biol.
**2009**, 58, 765–798. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Preziosi, L.; Tosin, A.R. Multiphase modeling of tumor growth and extra cellular matrix interaction: Mathematical tools and applications. J. Math. Biol.
**2019**, 58, 625–656. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Preziosi, L. (Ed.) Cancer Modelling and Simulation; Chapman & Hall/CRC Press: Boca Raton, FL, USA, 2003. [Google Scholar]
- Araujo, R.P.; McElwain, D.L.S. A history of the study of solid tumour growth: The contribution of mathematical modeling. Bull. Math. Biol.
**2004**, 66, 1039–1091. [Google Scholar] [CrossRef] [PubMed] - Astanin, S.; Preziosi, L. Multiphase models of tumour growth. In Selected Topics in Cancer Modelling: Genesis, Evolution, Immune Competition, and Therapy; Bellomo, N., Chaplain, M., de Angelis, E., Eds.; Birkhauser: Boston, MA, USA, 2008; pp. 223–254. [Google Scholar]
- Graziano, L.; Preziosi, L. Mechanics in tumour growth. In Modeling of Biological Materials; Mollica, F., Preziosi, L., Rajagopal, K.R., Eds.; Birkhauser: Boston, MA, USA, 2007; pp. 263–321. [Google Scholar]
- Roose, T.; Chapman, S.J.; Maini, P.K. Mathematical models of avascular tumor growth. SIAM Rev.
**2007**, 49, 179–208. [Google Scholar] [CrossRef] [Green Version] - Tracqui, P. Biophysical models of tumour growth. Rep. Prog. Phys.
**2009**, 72, 056701. [Google Scholar] [CrossRef] [Green Version] - Lowengrub, J.S.; Frieboes, H.B.; Jin, F.; Chuang, Y.L.; Li, X.; Macklin, P.; Wise, S.M.; Cristini, V. Nonlinear modelling of cancer: Bridging the gap between cells and tumours. Nonlinearity
**2010**, 23, R1–R9. [Google Scholar] [CrossRef] [Green Version] - Greenspan, H.P. On the growth and stabiligy of cell cultures and solid tumours. J. Theor. Biol.
**1976**, 56, 229–242. [Google Scholar] [CrossRef] - Adam, J.A. A simplified mathematical model of tumour growth. Math. Biosci.
**1986**, 81, 229–244. [Google Scholar] [CrossRef] - Chaplain, M.; Graziano, L.; Preziosi, L. Mathematical modelling of the loss of tissue compression responsiveness and its role in solid tumour development. Math. Med. Biol.
**2006**, 23, 197–229. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Byrne, H.M.; Chaplain, M.A.J. Modelling the role of cell–cell adhesion in the growth and developement of carcinoma. Math. Comput. Model.
**1996**, 24, 1–17. [Google Scholar] [CrossRef] - Armstrong, N.J.; Painter, K.J.; Sherratt, J.A. A continuum approach to modelling cell-cell adhesion. J. Theor. Biol.
**2006**, 243, 98–113. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Chaplain, M.A.J.; Lachowicz, M.; Szymanska, Z.; Wrzosek, D. Mathematical modelling of cancer invasion: The importance of cell–cell adhesion and cell-matrix adhesion. Math. Model. Methods Appl. Sci.
**2011**, 21, 719–743. [Google Scholar] [CrossRef] [Green Version] - Domschke, P.; Trucu, D.; Gerisch, A.; Chaplain, M.A.J. Mathematical modelling of cancer invasion: Implications of cell adhesion variability for tumour infiltrative growth patterns. J. Theor. Biol.
**2014**, 361, 41–60. [Google Scholar] [CrossRef] [Green Version] - Deakin, N.E.; Chaplain, M.A.J. Mathematical modelling of cancer invasion: The role of membrane-bound matrix metalloproteinases. Front. Oncol.
**2013**, 3, 70. [Google Scholar] [CrossRef] [Green Version] - Chaplain, M.A.J.; Lolas, G. Mathematical modelling of cancer cell invasion of tissue: The role of the urokinase plasminogen activation system. Math. Model. Methods Appl. Sci.
**2005**, 15, 1685–1734. [Google Scholar] [CrossRef] [Green Version] - Andasari, V.; Gerisch, A.; Lolas, G.; South, A.P.; Chaplain, M.A.J. Mathematical modeling of cancer cell invasion of tissue: Biological insight from mathematical analysis and computational simulation. J. Math. Biol.
**2011**, 63, 141–171. [Google Scholar] [CrossRef] - Ambrosi, D.; Preziosi, L. On the closure of mass balance models for tumour growth. Math. Models Methods Appl. Sci.
**2002**, 12, 737–754. [Google Scholar] [CrossRef] - Byrne, H.M. Modelling avascular tumor growth. In Cancer Modelling and Simulation; Preziosi, L., Ed.; Chapman & Hall/CRC Press: Boca Raton, FL, USA, 2003; pp. 75–120. [Google Scholar]
- Byrne, H.; Preziosi, L. Modelling solid tumour growth using the theory of mixtures. Math. Med. Biol.
**2003**, 20, 341–366. [Google Scholar] [CrossRef] - Chaplain, M.A.J.; Ganesh, M.; Graham, I.G. Spatio-temporal pattern formation on spherical surfaces: Numerical simulation and application to solid tumour growth. J. Math. Biol.
**2001**, 42, 387–423. [Google Scholar] [CrossRef] [PubMed] - Sherratt, J.A.; Chaplain, M.A.J. A new mathematical model for avascular tumour growth. J. Math. Biol.
**2001**, 43, 291–312. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Anderson, A.R.A.; Chaplain, M.A.J. Continuous and discrete mathematical models of tumour-induced angiogenesis. Bull. Math. Biol.
**1998**, 60, 857–899. [Google Scholar] [CrossRef] [PubMed] - Levine, H.A.; Pamuk, S.; Sleeman, B.D.; Nilsen-Hamilton, M. Mathematical modeling of capillary formation and development in tumor angiogenesis: Penetration into the stroma. Bull. Math. Biol.
**2001**, 63, 801–863. [Google Scholar] [CrossRef] - Levine, H.A.; Sleeman, B.D. Modelling tumour induced angiogenesis. In Cancer Modelling and Simulation; Preziosi, L., Ed.; Chapman & Hall/CRC Press: Boca Raton, FL, USA, 2003; pp. 147–184. [Google Scholar]
- McDougall, S.R.; Anderson, A.R.A.; Chaplain, M.A.J.; Sherrat, J.A. Mathematical modelling of flow through vascular networks: Implications for tumor induced angiogenesis and chemotherapy strategies. Bull. Math. Biol.
**2002**, 64, 673–702. [Google Scholar] [CrossRef] [PubMed] [Green Version] - McDougall, S.R.; Anderson, A.R.A.; Chaplain, M.A.J. Mathematical modelling of dynamic adaptive tumour-induced angiogenesis: Clinical implications and therapeutic targeting strategies. J. Theor. Biol.
**2006**, 241, 564–589. [Google Scholar] [CrossRef] - Folkman, J. Tumor angiogenesis. Adv. Cancer Res.
**1974**, 19, 331–358. [Google Scholar] [PubMed] - Folkman, J. The vascularization of tumors. Sci. Am.
**1976**, 234, 58–73. [Google Scholar] [CrossRef] - Folkman, J.; Klagsbrun, M. Angiogenic factors. Science
**1987**, 235, 442–447. [Google Scholar] [CrossRef] - Chaplain, M.A.J.; Anderson, A.R.A. Mathematical modelling of tissue invasion. In Cancer Modelling and Simulation; Preziosi, L., Ed.; Chapman & Hall/CRC Press: Boca Raton, FL, USA, 2003; pp. 269–298. [Google Scholar]
- Anderson, A.R.A.; Chaplain, M.A.J.; Newman, E.L.; Steele, R.J.C.; Thompson, A.M. Mathematical modeling of tumour invasion and metastasis. J. Theor. Med.
**2000**, 2, 129–154. [Google Scholar] [CrossRef] [Green Version] - Andreasen, P.A.; Kjøller, L.; Christensen, L.; Duffy, M.J. The urokinase-type plasminogen activator system in cancer metastasis: A review. Int. J. Cancer
**1997**, 72, 1–22. [Google Scholar] [CrossRef] - Andreasen, P.A.; Egelund, R.; Petersen, H.H. The plasminogen activation system in tumor growth, invasion, and metastasis. Cell Mol. Life Sci.
**2000**, 57, 25–40. [Google Scholar] [CrossRef] - Simpson, A.J.; Booth, N.A.; Moore, N.R.; Bennett, B. Distribution of plasminogen activator inhibitor (PAI-1) in tissue. J. Clin. Pathol.
**1991**, 44, 139–143. [Google Scholar] [CrossRef] [Green Version] - Wiman, B.; Almquist, A.; Sigurdardottir, O.; Lindahl, T. Plasminogen activator inhibitor 1 (PAI) is bound to vitronectin in plasma. FEBS Lett.
**1988**, 242, 125–128. [Google Scholar] [CrossRef] [Green Version] - Stefanson, S.; Lawrence, D.A. The serpin PAI-1 inhibits cell migration by blocking integrin α
_{v}β_{3}binding to vitronectin. Nature**1996**, 383, 441–443. [Google Scholar] [CrossRef] [PubMed] - Waltz, D.A.; Natkin, L.R.; Fujita, R.M.; Wei, Y.; Chapman, H.A. Plasmin and plasminogen activator inhibitor type 1 promote cellular motility by regulating the interaction between the urokinase receptor and vitronectin. J. Clin. Investig.
**1997**, 100, 58–67. [Google Scholar] [CrossRef] [Green Version] - Frieboes, H.B.; Jin, F.; Chuang, Y.L.; Wise, S.M.; Lowengrub, J.S.; Cristini, V. Three-dimensional multispecies nonlinear tumor growth-ii: Tumor invasion and angiogenesis. J. Theor. Biol.
**2010**, 264, 1254–1278. [Google Scholar] [CrossRef] [Green Version] - Mascheroni, P.; Carfagna, M.; Grillo, A.; Boso, D.P.; Schrefler, B.A. An avascular tumor growth model based on porous media mechanics and evolving natural states. Math. Mech. Solids
**2018**, 23, 686–712. [Google Scholar] [CrossRef] - Wise, S.M.; Lowengrub, J.S.; Cristini, V. An adaptive multigrid algorithm for simulating solid tumour growth using mixture models. Math. Comput. Model.
**2011**, 53, 1–20. [Google Scholar] [CrossRef] - Frieboes, H.B.; Zheng, X.; Sun, C.H.; Tromberg, B.; Gatenby, R.; Cristini, V. An integrated computational/experimental model of tumor invasion. Cancer Res.
**2006**, 66, 1597–1604. [Google Scholar] [CrossRef] [Green Version] - Macklin, P.; Lowengrub, J. Nonlinear simulation of the effect of microenvironment on tumor growth. J. Theor. Biol.
**2007**, 245, 677–704. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Zheng, X.; Wise, S.; Cristini, V. Nonlinear simulation of tumor necrosis, neo-vascularization and tissue invasion via an adaptive finite-element/level-set method. Bull. Math. Biol.
**2005**, 67, 211–259. [Google Scholar] [CrossRef] [PubMed] - Cristini, V.; Li, X.; Lowengrub, J.S.; Wise, S.M. Nonlinear simulations of solid tumor growth using a mixture model: Invasion and branching. J. Math. Biol.
**2009**, 58, 723–763. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Hubbard, M.E.; Byrne, H.M. Multiphase modeling of vascular tumour growth in two spatial dimensions. J. Theor. Biol.
**2013**, 316, 70–89. [Google Scholar] [CrossRef] [PubMed] - Peng, L.; Trucu, D.; Lin, P.; Thompson, A.; Chaplain, M.A.J. A multiscale mathematical model of tumour invasive growth. Bull. Math. Biol.
**2017**, 79, 389–429. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Amoddeo, A. Adaptive grid modelling for cancer cells in the early stage of invasion. Comput. Math. Appl.
**2015**, 69, 610–619. [Google Scholar] [CrossRef] - Amoddeo, A. Moving mesh partial differential equations modelling to describe oxygen induced effects on avascular tumour growth. Cogent Phys.
**2015**, 2, 1050080. [Google Scholar] [CrossRef] - Amoddeo, A. Oxygen induced effects on avascular tumour growth: A preliminary simulation using an adaptive grid algorithm. J. Phys. Conf. Ser.
**2015**, 633, 012088. [Google Scholar] [CrossRef] - Amoddeo, A. A moving mesh study for diffusion induced effects in avascular tumour growth. Comput. Math. Appl.
**2018**, 75, 2508–2519. [Google Scholar] [CrossRef] - Amoddeo, A. Modelling avascular tumor growth: Approach with an adaptive grid numerical technique. J. Multiscale Model.
**2018**, 9, 1840002. [Google Scholar] [CrossRef] - Zienkiewicz, O.C.; Taylor, R.L. The Finite Element Method; Butterworth–Heinemann: Oxford, UK, 2002. [Google Scholar]
- Chun, M.H. Plasmin induces the formation of multicellular spheroids of breast cancer cells. Cancer Lett.
**1997**, 117, 51–56. [Google Scholar] [CrossRef] - Praus, M.; Collen, D.; Gerard, R.D. Both u-PA inhibition and vitronectin binding by plasminogen activator inhibitor 1 regulate HT1080 fibrosarcoma cell metastasis. Int. J. Cancer
**2002**, 102, 584–591. [Google Scholar] [CrossRef] [PubMed]

**Figure 1.**Visual overview schematizing the interactions among the species inside the simulated domain, imposing zero-flux boundary conditions. Red dashed lines refer to indirect cancer cells proliferation due to plasminogen activator inhibitor tpe-1 (PAI-1)/vitronectin (VN) binding; green dashed lines refer to indirect plasmin deactivation from PAI-1/urokinase plasminogen activator (uPA) binding.

**Figure 2.**Snapshots of cancer cell concentrations obtained in the simulated domain imposing ϕ

_{51}= 0. The non-dimensional time steps are t = 40, 60, 80, and 100, while ϕ

_{14}= 0, 0.275, and 0.55; the other parameters are as in Table 1. The concentration is linearly mapped on a color scale between the blue and yellow colors.

**Figure 3.**Snapshots of VN concentrations obtained in the simulated domain imposing ϕ

_{51}= 0. The non-dimensional time steps are t = 40, 60, 80, and 100, while ϕ

_{14}= 0, 0.275, and 0.55; the other parameters are as in Table 1. The concentration is linearly mapped on a color scale between the blue and yellow colors.

**Figure 4.**Snapshots of plasmin concentrations obtained in the simulated domain imposing ϕ

_{51}= 0. The non-dimensional time steps are t = 40, 60, 80, and 100, while ϕ

_{14}= 0, 0.275, and 0.55; the other parameters are as in Table 1. The concentration is linearly mapped on a color scale between the blue and yellow colors.

**Figure 5.**Snapshots of cancer cell concentrations obtained in the simulated domain imposing ϕ

_{51}= 0.375. The non-dimensional time steps are t = 40, 60, 80, and 100, while ϕ

_{14}= 0, 0.275, and 0.55; the other parameters are as in Table 1. The concentration is linearly mapped on a color scale between the blue and yellow colors.

**Figure 6.**Snapshots of VN concentrations obtained in the simulated domain imposing ϕ

_{51}= 0.375. The non-dimensional time steps are t = 40, 60, 80, and 100, while ϕ

_{14}= 0, 0.275, and 0.55; the other parameters are as in Table 1. The concentration is linearly mapped on a color scale between the blue and yellow colors.

**Figure 7.**Snapshots of plasmin concentration obtained in the simulated domain imposing ϕ

_{51}= 0.375. The non-dimensional time steps are t = 40, 60, 80, and 100, while ϕ

_{14}= 0, 0.275, and 0.55; the other parameters are as in Table 1. The concentration is linearly mapped on a color scale between the blue and yellow colors.

**Figure 8.**Snapshots of cancer cell concentrations obtained in the simulated domain imposing ϕ

_{51}= 0.75. The non-dimensional time steps are t = 40, 60, 80, and 100, while ϕ

_{14}= 0, 0.275, and 0.55; the other parameters are as in Table 1. The concentration is linearly mapped on a color scale between the blue and yellow colors.

**Figure 9.**Snapshots of VN concentrations obtained in the simulated domain imposing ϕ

_{51}= 0.75. The non-dimensional time steps are t = 40, 60, 80, and 100, while ϕ

_{14}= 0, 0.275, and 0.55; the other parameters are as in Table 1. The concentration is linearly mapped on a color scale between the blue and yellow colors.

**Figure 10.**Snapshots of plasmin concentrations obtained in the simulated domain imposing ϕ

_{51}= 0.75. The non-dimensional time steps are t = 40, 60, 80, and 100, while ϕ

_{14}= 0, 0.275, and 0.55; the other parameters are as in Table 1. The concentration is linearly mapped on a color scale between the blue and yellow colors.

Parameter | Units | Non-Dimensional Parameter | Value | Description |
---|---|---|---|---|

D_{c} | cm^{2}s^{−1} | D_{c}/D | 3.5 × 10^{−4} | Cancer cell diffusion coefficient |

D_{u} | cm^{2}s^{−1} | D_{u}/D | 2.5 × 10^{−3} | uPA ^{1} diffusion coefficient |

D_{p} | cm^{2}s^{−1} | D_{p}/D | 3.5 × 10^{−3} | PAI – 1 ^{2} diffusion coefficient |

D_{m} | cm^{2}s^{−1} | D_{m}/D | 4.91 × 10^{−3} | Plasmin diffusion coefficient |

χ_{u} | cm^{2}s^{−1}nM^{−1} | u_{0}D^{−1}χ_{u} | 3.05 × 10^{−2} | uPA chemotactic coefficient |

χ_{p} | cm^{2}s^{−1}nM^{−1} | p_{0}D^{−1}χ_{p} | 3.75 × 10^{−2} | PAI -1 chemotactic coefficient |

χ_{v} | cm^{2}s^{−1}nM^{−1} | v_{0}D^{−1}χ_{v} | 2.85 × 10^{−2} | VN ^{3} haptotactic coefficient |

μ_{1} | s^{−1} | τμ_{1} | 0.25 | Cancer cell proliferation rate |

μ_{2} | s^{−1} | τμ_{2} | 0.15 | ECM ^{4} proliferation rate |

α_{31} | cell^{−1}cm^{3}s^{−1}nM | τc_{0}u_{0}^{−1}α_{31} | 0.215 | uPA production rate |

α_{41} | s^{−1} | τm_{0}p_{0}^{−1}α_{41} | 0.5 | PAI - 1 production rate |

δ | s^{−1}nM^{−1} | τm_{0}δ | 8.15 | VN degradation rate from interaction with plasmin |

ϕ_{14} | cell cm^{−3}s^{−1}nM^{−2} | τp_{0}v_{0}c_{0}^{−1}ϕ_{14} | 0, 0.275, 0.55 | Cancer cell proliferation rate from PAI-1/VN interaction |

ϕ_{21} | s^{−1}nM^{−1} | τu_{0}p_{0}v_{0}^{−1}ϕ_{21} | 0.75 | VN production rate from uPA/PAI-1 interaction |

ϕ_{22} | s^{−1}nM^{−1} | τp_{0}ϕ_{22} | 0.55 | VN neutralization rate from interaction with PAI-1 |

ϕ_{31} | s^{−1}nM^{−1} | τp_{0}ϕ_{31} | 0.75 | uPA inhibition rate from interaction with PAI-1 |

ϕ_{33} | cell^{−1}cm^{3}s^{−1} | τc_{0}ϕ_{33} | 0.3 | uPA degradation rate from interaction with uPAR |

ϕ_{41} | s^{−1}nM^{−1} | τu_{0}ϕ_{41} | 0.75 | PAI-1 degradation rate from interaction with uPA |

ϕ_{42} | s^{−1}nM^{−1} | τv_{0}ϕ_{42} | 0.55 | PAI-1 degradation rate from interaction with VN |

ϕ_{51} | s^{−1}nM^{−1} | τp_{0}u_{0}m_{0}^{−1}ϕ_{51} | 0, 0.375, 0.75 | Plasmin deactivation rate from PAI-1/uPA interaction |

ϕ_{52} | s^{−1}nM^{−1} | τp_{0}v_{0}m_{0}^{−1}ϕ_{52} | 0.11 | Plasmin production rate from PAI-1/VN interaction |

ϕ_{53} | cell^{−1}cm^{3}s^{−1} | τu_{0}c_{0}m_{0}^{−1}ϕ_{53} | 0.75 | Plasmin production rate from cancer cell/uPA interaction |

ϕ_{54} | s^{−1} | τϕ_{54} | 0.5 | Plasmin degradation rate |

^{1}urokinase plasminogen activator;

^{2}plasminogen activator inhibitor type-1;

^{3}vitronectin;

^{4}extra-cellular matrix.

© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Amoddeo, A.
Indirect Contributions to Tumor Dynamics in the First Stage of the Avascular Phase. *Symmetry* **2020**, *12*, 1546.
https://doi.org/10.3390/sym12091546

**AMA Style**

Amoddeo A.
Indirect Contributions to Tumor Dynamics in the First Stage of the Avascular Phase. *Symmetry*. 2020; 12(9):1546.
https://doi.org/10.3390/sym12091546

**Chicago/Turabian Style**

Amoddeo, Antonino.
2020. "Indirect Contributions to Tumor Dynamics in the First Stage of the Avascular Phase" *Symmetry* 12, no. 9: 1546.
https://doi.org/10.3390/sym12091546