Currently, traditional 2D and 3D cancer models are the most used strategies for the study and the development of new drugs or pharmaceutical therapies. The lack of realistic modeling leads to low predictivity of effects and safety, and the high costs and time-consuming processes make them not always available and usable. Interest in microfluidic devices, in which it is possible to recreate perfusion and shear stresses, is driven by the need of a more physiological representation of the sample to be studied. Moreover, traditional sensors and characterization techniques provide information about cellular response, but often they are destructive for the tissue, incompatible with the device, expensive and give information only at the end of the process.