Immune checkpoint blockade (ICB) has shown unprecedented clinical success in treatment of cancer. However, not all patients show adequate response, and the treatment can lead to a broad range of adverse effects. Therefore, early identification of potential responders to therapy, using non-invasive means, is a critical challenge for improving ICB. Herein, we engineered anti-Programmed Death Ligand 1 (aPDL1) nanoparticles with enhanced ICB immunotherapy efficacy. Using a mouse model for colon cancer, we show that the nanoparticles accumulated, penetrated and efficiently prevented tumor growth. Moreover, we found a direct correlation between the amount of nanoparticle accumulation within the tumor at 48 hours, as determined by CT, and the therapeutic response. This enabled subject stratification as potential responders or non-responders, at an early time point. Thus, by integrating prognostic and ICB-based therapeutic functions into one nanoparticle, we obtained a straightforward approach for potential imageguided stratification of cancer patient subpopulations.