Data driven modeling of pseudopalisade pattern formation.

Pseudopalisading is an interesting phenomenon where cancer cells arrange themselves to form a dense garland-like pattern. Unlike the palisade structure, a similar type of pattern first observed in schwannomas by pathologist J.J. Verocay (Wippold et al. in AJNR Am J Neuroradiol 27(10):2037-2041, 2006), pseudopalisades are less organized and associated with a necrotic region at their core. These structures are mainly found in glioblastoma (GBM), a grade IV brain tumor, and provide a way to assess the aggressiveness of the tumor. Identification of the exact bio-mechanism responsible for the formation of pseudopalisades is a difficult task, mainly because pseudopalisades seem to be a consequence of complex nonlinear dynamics within the tumor. In this paper we propose a data-driven methodology to gain insight into the formation of different types of pseudopalisade structures. To this end, we start from a state of the art macroscopic model for the dynamics of GBM, that is coupled with the dynamics of extracellular pH, and formulate a terminal value optimal control problem. Thus, given a specific, observed pseudopalisade pattern, we determine the evolution of parameters (bio-mechanisms) that are responsible for its emergence. Random histological images exhibiting pseudopalisade-like structures are chosen to serve as target pattern. Having identified the optimal model parameters that generate the specified target pattern, we then formulate two different types of pattern counteracting ansatzes in order to determine possible ways to impair or obstruct the process of pseudopalisade formation. This provides the basis for designing active or live control of malignant GBM. Furthermore, we also provide a simple, yet insightful, mechanism to synthesize new pseudopalisade patterns by linearly combining the optimal model parameters responsible for generating different known target patterns. This particularly provides a hint that complex pseudopalisade patterns could be synthesized by a linear combination of parameters responsible for generating simple patterns. Going even further, we ask ourselves if complex therapy approaches can be conceived, such that some linear combination thereof is able to reverse or disrupt simple pseudopalisade patterns; this is investigated with the help of numerical simulations.

Feasibility and clinical usefulness of modelling glioblastoma migration in adjuvant radiotherapy.

Glioblastoma (GBM) is one of the most common primary brain tumours in adults, with a dismal prognosis despite aggressive multimodality treatment by a combination of surgery and adjuvant radiochemotherapy. A detailed knowledge of the spreading of glioma cells in the brain might allow for more targeted escalated radiotherapy, aiming to reduce locoregional relapse. Recent years have seen the development of a large variety of mathematical modelling approaches to predict glioma migration. The aim of this study is hence to evaluate the clinical applicability of a detailed micro- and meso-scale mathematical model in radiotherapy. First and foremost, a clinical workflow is established, in which the tumour is automatically segmented as input data and then followed in time mathematically based on the diffusion tensor imaging data. The influence of several free model parameters is individually evaluated, then the full model is retrospectively validated for a collective of 3 GBM patients treated at our institution by varying the most important model parameters to achieve optimum agreement with the tumour development during follow-up. Agreement of the model predictions with the real tumour growth as defined by manual contouring based on the follow-up MRI images is analyzed using the dice coefficient. The tumour evolution over 103-212 days follow-up could be predicted by the model with a dice coefficient better than 60% for all three patients. In all cases, the final tumour volume was overestimated by the model by a factor between 1.05 and 1.47. To evaluate the quality of the agreement between the model predictions and the ground truth, we must keep in mind that our gold standard relies on a single observer's (CB) manually-delineated tumour contours. We therefore decided to add a short validation of the stability and reliability of these contours by an inter-observer analysis including three other experienced radiation oncologists from our department. In total, a dice coefficient between 63% and 89% is achieved between the four different observers. Compared with this value, the model predictions (62-66%) perform reasonably well, given the fact that these tumour volumes were created based on the pre-operative segmentation and DTI.

Multiscale modeling of glioma pseudopalisades: contributions from the tumor microenvironment.

Gliomas are primary brain tumors with a high invasive potential and infiltrative spread. Among them, glioblastoma multiforme (GBM) exhibits microvascular hyperplasia and pronounced necrosis triggered by hypoxia. Histological samples showing garland-like hypercellular structures (so-called pseudopalisades) centered around the occlusion site of a capillary are typical for GBM and hint on poor prognosis of patient survival. We propose a multiscale modeling approach in the kinetic theory of active particles framework and deduce by an upscaling process a reaction-diffusion model with repellent pH-taxis. We prove existence of a unique global bounded classical solution for a version of the obtained macroscopic system and investigate the asymptotic behavior of the solution. Moreover, we study two different types of scaling and compare the behavior of the obtained macroscopic PDEs by way of simulations. These show that patterns (not necessarily of Turing type), including pseudopalisades, can be formed for some parameter ranges, in accordance with the tumor grade. This is true when the PDEs are obtained via parabolic scaling (undirected tissue), while no such patterns are observed for the PDEs arising by a hyperbolic limit (directed tissue). This suggests that brain tissue might be undirected - at least as far as glioma migration is concerned. We also investigate two different ways of including cell level descriptions of response to hypoxia and the way they are related .

Nonlocal and local models for taxis in cell migration: a rigorous limit procedure.

A rigorous limit procedure is presented which links nonlocal models involving adhesion or nonlocal chemotaxis to their local counterparts featuring haptotaxis and classical chemotaxis, respectively. It relies on a novel reformulation of the involved nonlocalities in terms of integral operators applied directly to the gradients of signal-dependent quantities. The proposed approach handles both model types in a unified way and extends the previous mathematical framework to settings that allow for general solution-dependent coefficient functions. The previous forms of nonlocal operators are compared with the new ones introduced in this paper and the advantages of the latter are highlighted by concrete examples. Numerical simulations in 1D provide an illustration of some of the theoretical findings.

Mathematical models for cell migration: a non-local perspective.

We provide a review of recent advancements in non-local continuous models for migration, mainly from the perspective of its involvement in embryonal development and cancer invasion. Particular emphasis is placed on spatial non-locality occurring in advection terms, used to characterize a cell's motility bias according to its interactions with other cellular and acellular components in its vicinity (e.g. cell-cell and cell-tissue adhesions, non-local chemotaxis), but we also briefly address spatially non-local source terms. Following a short introduction and description of applications, we give a systematic classification of available PDE models with respect to the type of featured non-localities and review some of the mathematical challenges arising from such models, with a focus on analytical aspects. This article is part of the theme issue 'Multi-scale analysis and modelling of collective migration in biological systems'.

Modeling of pH regulation in tumor cells: Direct interaction between proton-coupled lactate transporters and cancer-associated carbonic anhydrase.

The most aggressive tumor cells, which often reside in a hypoxic environment, can release vast amounts of lactate and protons via monocarboxylate transporters (MCTs). This additional proton efflux exacerbates extracellular acidification and supports the formation of a hostile environment. In the present study we propose a novel, data-based model for this proton-coupled lactate transport in cancer cells. The mathematical settings involve systems coupling nonlinear ordinary and stochastic differential equations describing the dynamics of intra- and extracellular proton and lactate concentrations. The data involve time series of intracellular proton concentrations of normoxic and hypoxic MCF-7 breast cancer cells. The good agreement of our final model with the data suggests the existence of proton pools near the cell membrane, which can be controlled by intracellular and extracellular carbonic anhydrases to drive proton-coupled lactate transport across the plasma membrane of hypoxic cancer cells.

A multiscale model for glioma spread including cell-tissue interactions and proliferation.

Glioma is a broad class of brain and spinal cord tumors arising from glia cells, which are the main brain cells that can develop into neoplasms. They are highly invasive and lead to irregular tumor margins which are not precisely identifiable by medical imaging, thus rendering a precise enough resection very difficult. The understanding of glioma spread patterns is hence essential for both radiological therapy as well as surgical treatment. In this paper we propose a multiscale model for glioma growth including interactions of the cells with the underlying tissue network, along with proliferative effects. Our current accounting for two subpopulations of cells to accomodate proliferation according to the go-or-grow dichtomoty is an extension of the setting in [16]. As in that paper, we assume that cancer cells use neuronal fiber tracts as invasive pathways. Hence, the individual structure of brain tissue seems to be decisive for the tumor spread. Diffusion tensor imaging (DTI) is able to provide such information, thus opening the way for patient specific modeling of glioma invasion. Starting from a multiscale model involving subcellular (microscopic) and individual (mesoscale) cell dynamics, we perform a parabolic scaling to obtain an approximating reaction-diffusion-transport equation on the macroscale of the tumor cell population. Numerical simulations based on DTI data are carried out in order to assess the performance of our modeling approach.

Effective equations for anisotropic glioma spread with proliferation: a multiscale approach and comparisons with previous settings.

Glioma is a common type of primary brain tumour, with a strongly invasive potential, often exhibiting non-uniform, highly irregular growth. This makes it difficult to assess the degree of extent of the tumour, hence bringing about a supplementary challenge for the treatment. It is therefore necessary to understand the migratory behaviour of glioma in greater detail. In this paper, we propose a multiscale model for glioma growth and migration. Our model couples the microscale dynamics (reduced to the binding of surface receptors to the surrounding tissue) with a kinetic transport equation for the cell density on the mesoscopic level of individual cells. On the latter scale, we also include the proliferation of tumour cells via effects of interaction with the tissue. An adequate parabolic scaling yields a convection-diffusion-reaction equation, for which the coefficients can be explicitly determined from the information about the tissue obtained by diffusion tensor imaging (DTI). Numerical simulations relying on DTI measurements confirm the biological findings that glioma spread along white matter tracts.

Glioma follow white matter tracts: a multiscale DTI-based model.

Gliomas are a class of rarely curable tumors arising from abnormal glia cells in the human brain. The understanding of glioma spreading patterns is essential for both radiological therapy as well as surgical treatment. Diffusion tensor imaging (DTI) allows to infer the white matter fibre structure of the brain in a noninvasive way. Painter and Hillen (J Theor Biol 323:25-39, 2013) used a kinetic partial differential equation to include DTI data into a class of anisotropic diffusion models for glioma spread. Here we extend this model to explicitly include adhesion mechanisms between glioma cells and the extracellular matrix components which are associated to white matter tracts. The mathematical modelling follows the multiscale approach proposed by Kelkel and Surulescu (Math Models Methods Appl Sci 23(3), 2012). We use scaling arguments to deduce a macroscopic advection-diffusion model for this process. The tumor diffusion tensor and the tumor drift velocity depend on both, the directions of the white matter tracts as well as the binding dynamics of the adhesion molecules. The advanced computational platform DUNE enables us to accurately solve our macroscopic model. It turns out that the inclusion of cell binding dynamics on the microlevel is an important factor to explain finger-like spread of glioma.

A minimal mathematical model for the initial molecular interactions of death receptor signalling.

Tumor necrosis factor (TNF) is the name giving member of a large cytokine family mirrored by a respective cell membrane receptor super family. TNF itself is a strong proinflammatory regulator of the innate immune system, but has been also recognized as a major factor in progression of autoimmune diseases. A subgroup of the TNF ligand family, including TNF, signals via so-called death receptors, capable to induce a major form of programmed cell death, called apoptosis. Typical for most members of the whole family, death ligands form homotrimeric proteins, capable to bind up to three of their respective receptor molecules. But also unligated receptors occur on the cell surface as homomultimers due to a homophilic interaction domain. Based on these two interaction motivs (ligand/receptor and receptor/receptor) formation of large ligand/receptor clusters can be postulated which have been also observed experimentally. We use here a mass action kinetics approach to establish an ordinary differential equations model describing the dynamics of primary ligand/receptor complex formation as a basis for further clustering on the cell membrane. Based on available experimental data we develop our model in a way that not only ligand/receptor, but also homophilic receptor interaction is encompassed. The model allows formation of two distict primary ligand/receptor complexes in a ligand concentration dependent manner. At extremely high ligand concentrations the system is dominated by ligated receptor homodimers.

Modeling and simulation of some cell dispersion problems by a nonparametric method.

Starting from the classical descriptions of cell motion we propose some ways to enhance the realism of modeling and to account for interesting features like allowing for a random switching between biased and unbiased motion or avoiding a set of obstacles. For this complex behavior of the cell population we propose new models and also provide a way to numerically assess the macroscopic densities of interest upon using a nonparametric estimation technique. Up to our knowledge, this is the only method able to numerically handle the entire complexity of such settings.

On some models for cancer cell migration through tissue networks.

We propose some models allowing to account for relevant processes at the various scales of cancer cell migration through tissue, ranging from the receptor dynamics on the cell surface over degradation of tissue fibers by protease and soluble ligand production towards the behavior of the entire cell population. For a genuinely mesoscopic version of these models we also provide a result on the local existence and uniqueness of a solution for all biologically relevant space dimensions.

On a stochastic reaction-diffusion system modeling pattern formation on seashells.

Starting from the Gierer-Meinhardt setting, we propose a stochastic model to characterize pattern formation on seashells under the influence of random space-time fluctuations. We prove the existence of a positive solution for the resulting system and perform numerical simulations in order to assess the behavior of the solution in comparison with the deterministic approach.