A mathematical model describing the time evolution of a Cancer cell population in Chronic Myelogenous Leukemia (CML) is extended: a function modeling the treatment approach of the disease is taken into account, allowing the investigation of the most appropriate cure protocol. Because CML is believed to arise in the hematopoietic stem cell (HSC) compartment, the study in this paper starts from a delay differential system modeling CML in the stem cell compartment. Optimal control is applied to this two-dimensional system with a molecular targeted therapy, such as imatinib. The objective in this work is to minimize the cancer cell population and the negative effects of the medication to the healthy cells of a certain patient. In order to solve the delayed optimal control problem, the system is discretized and its behavior is investigated for given sets of parameters. The obtained results are illustrated by numerical simulations and discussed in view of the optimal treatment approach.