In self-triggered network control systems, next sampling or control time instant is computed based on the last measurement of system states. In this paper, inspired by self-triggered strategy, a new approach to closed loop chemotherapeutic dose regimen design is introduced. The proposed scheme meets clinical constraints such as discontinuous drug administration and sparse measurement of cancer cells. Theoretical outcomes confirm that the proposed therapy protocol results in cancer remission. The ultimate bound of cancer cells can be set as desired. The proposed therapy protocol is tested on a set of 5 virtual patients. The in-silico results show the efficacy and robustness of the proposed strategy.