Data-driven modeling of combination therapy in oncology

T. Cardilin. Licentiate thesis, Chalmers University of Technology and University of Gothenburg, supervisors M. Jirstrand, T. Lundh and J. Gabrielsson, September 2016.


This thesis contains two manuscripts: Tumor Static Concentration Curves in Combination Therapy and Extending the Tumor Static Concentration Curve to Exposure – A Combination Therapy Example with Radiation Therapy. There is also an introductory chapter presenting some basic facts necessary to understand the appended manuscripts. The manuscripts share the common goal of developing model-based data-driven tools and techniques to quantitatively assess the effectiveness of anticancer combinations. The first paper presents a dynamical systems model for combination therapy with the anticancer drugs cetuximab and cisplatin. Using a mixed-effects approach the model is shown to adequately describe a preclinical dataset. The model is then analyzed by introducing the Tumor Static Concentration (TSC) curve, a curve of cetuximab-cisplatin concentration pairs all of which, if maintained, result in tumor stasis. The TSC analysis reveals a modest gain from combining the compounds. The variability of the TSC curve across the population is also explored. In the second paper we develop a dynamical systems model for combination therapy with ionizing radiation and a test compound. For this combination we introduce an extension of the TSC curve called the (average) Tumor Static Exposure (TSE) curve based on average, as opposed to pointwise, tumor stasis. The TSE analysis for combinations of radiation and the test compound demonstrates a large synergistic effect.


Photo credits: Nic McPhee