A novel numerical and artificial intelligence based approach to study anti-angiogenic drugs: Endostatin

Abstract

Angiogenesis is known as the key factor in vascular tumor growth. In addition to furnishing the tumor with fresh nutrients to survive and spread to the other tissues, tumor-induced vasculature is tortuous and contains leaky vessels. Therefore, anti-angiogenesis is introduced as a way to suppress tumor growth and preclude metastases. Secondly, method of normalization of tortuous tumor-induced vasculature is being used to ease drug delivery to the tumor site. And thirdly, combined therapies of chemotherapy and radiotherapy with anti-angiogenesis agents can help achieve better outcomes in cancer treatment. One endogenous anti-angiogenesis agent that is being used in the aforementioned treatment strategies is known to be endostatin. In this research, a novel formulation is proposed and a computer code is developed to study anti-angiogenesis effects of endostatin. It is shown that endostatin as an endogenous agent can suppress angiogenesis and normalize tumor-induced vasculature which can ease drug delivery to the tumor site. By increasing endostatin concentration to 5 times of its natural concentration in the blood plasma of cancer patients, angiogenesis could be tackled and hindered. Finally, in order to propose a general and simple formulation to predict final microvessel density at different circumstances, a Generalized Regression Neural Network (GRNN) is established. The results of GRNN show that it is able to predict the microvascular density with 87% accuracy. With the aim of GRNN formulation, scientists can observe the vasculature at any desired conditions to get an insight on optimum time for the combined treatment.