Section: New Results

Cancer modeling

• Patient specific simulation for lung metastases

  The calibration process has tremendously improved by a deep study of the model and its parameter space. Work is ongoing to validate the whole process on a retrospective study of 30 patients. A prototype is being built for our collaborators at Institut Bergonié to use in their clinical routine. The same strategy has been applied to meningiomas in the last year of the post-doc of Julie Joie within the IRL MONICA with a retrospectuve study on 10 patients.

• Modelling of the response to targeted therapies for liver metastasis of a gist : 2 clinical cases with a long term longitudinal follow-up with CT-scans. We are able to fit the volume of the lesion but also the the texture of the image, that is the ratio between necrotic tissues and proliferative ones. See [82] .

• Tumor growth model for ductal carcinoma: from in situ phase to stroma invasion. See [71] .

• Permeable and conducting states of membrane submitted to electric pulse: non-linear PDE model, 2D and 3D code in C++.

• Free boundary value model for invadopodia and migration of cell developed in collaboration with Osaka University and Tokyo University of Sciences.

• Endothelial cell migration on polymers: agent based model. Paper accepted in DCDS-B.

• A. Peretti started her PhD on the modeling of the heterogenity on renal cancer.

• Benjamin Taton started a post-doc on the modeling of the renal function through perfusion MRI. B. Taton is a MD.

• Th. Michel obtained some mathematical properties on the system of PDEs used for the modelling of GIST metastases.

• Models for preclinical studies

  • Mathematical ODE models of tumor volume kinetics in mice (collaboration with the Center of Cancer and Systems Biology, Boston, USA and J. Ebos, Roswell Park Cancer Institute, Buffalo, USA).

    Rational and quantitative evaluation of the predictive and descriptive power of the majority of the classical ODE models for tumor growth against data from two distinct experimental systems [57] . One of the major finding was the huge improvement of the predictive properties when using the population a priori information on the distribution of the parameters.

  • Mathematical model for data of preclinical metastatic burden dynamics and clinical data of metastatic relapse probability of breast cancer (collaboration with J. Ebos, Roswell Park Cancer Institute, Buffalo, USA).

    Validation of the descriptive and predictive ability of a simple and minimally parameterized model. The major finding resulting from the modeling analysis was the quantification of the impact of surgery on survival improvement (highly nonlinear), which suggests a threshold primary tumor size for efficacy of the surgery in terms of preventing metastatic recurrence. A publication is in preparation.

  • Effect of anti-cancer therapies in preclinical experiments

    • Evaluation of several models (several already published but also new ones) for the effect of anti-angiogenic drugs (recent anti-cancer drugs that target the tumor vasculature rather than the cancer cells themselves) on tumor growth, based on statistical parameter estimation methods on experimental data (collaboration with J. Ebos, Roswell Park Cancer Institute, Buffalo, USA). The main finding was one model that was able to both describe the effect of the drug (Sunitinib) and predict the effect when changing the scheduling. See [66] .

    • Effect of the sequence of administration between cytotoxic and anti-angiogenic drugs (collaboration with J. Ciccolini and D. Barbolosi, SMARTc, Inserm, Marseille, Fr). See [84] .

• Theoretical cancer biology

  • Theories of metastatic initiation (collaboration with A. Bikfalvi, LAMC, Inserm and the RMSB, CNRS in Bordeaux, Fr).

    Confrontation of theories and experimental data challenged the classical view of metastatic establishment and growth and suggested that tumors could merge in initial phases. Quantitative impact of the merging was studied using a dedicated and properly calibrated spatial model.

  • Tumor-tumor distant interactions (collaboration with the Center of Cancer and Systems Biology, Boston, USA).

    Statistical and modeling analysis of experimental data for two tumors implanted in one organism.