Sensitivity analysis for a mathematical model of tumor-immune interactions

A mathematical model, in the form of a system of nonlinear ordinary differential equations, has been utilized to investigate the interaction between immune cells and tumor cells. Immune cells included are natural killer cells, circulating lymphocytes, CD8+T cells, CD4+T cells and cytokines. The dual role of the CD4+T cells in acti-vating CD8+T cells and in killing the tumor via secretion of cytokines is represented in the model. Sensitivity analysis is performed, for two sets of human data, to identify the most effective body parameters in eliminating tumor cell population. Both numerical sensitivity coefficient method and sensitivity function method are implemented, and the results are compared. While the first method identifies the most effective parameters at a specific instant of time, like many published works, the second method recognizes these parameters over a wide time interval. The required order of the most effective parameters is identified, so this sensitivity analysis answers the question: Which parameter can be changed to get the largest effect on the tumor size? The obtained results provide a valuable tool to identify the parameters that would be increased or decreased prior to starting a treatment. © 2021, Politechnica University of Bucharest. All rights reserved.