Granzyme B PET imaging of immune-mediated tumor killing as a tool for understanding immunotherapy response

BACKGROUND: Cancer immunotherapy research is expanding to include a more robust understanding of the mechanisms of treatment response and resistance. Identification of drivers of pro-tumor and anti-tumor immunity during treatment offers new strategies for effective alternative or combination immunotherapies. Currently, tissue or blood samples are collected and analyzed, then dichotomized based on clinical end points that may occur months or years after tissue is collected. While overall survival is ultimately the desired clinical outcome, this dichotomization fails to incorporate the nuances that may occur during an anti-tumor response. By failing to directly measure immune activation at the time of sampling, tumors may be misclassified and potentially obscure important biological information. Non-invasive techniques, such as positron emission tomography (PET), allow for global and quantitative measurements of cancer specific processes and are widely used clinically to help manage disease. METHODS: We have previously developed a novel PET agent that can non-invasively quantify granzyme B release in tumors and have demonstrated its ability to predict response to checkpoint inhibitor therapy in multiple murine models of cancer. Here, we used the quantitative measurement of granzyme B release as a direct and time-matched marker of immune cell activation in order to determine immune cell types and cytokines that correlate with effective checkpoint inhibitor therapy in both tumors and tumor-draining lymph nodes. RESULTS: Through PET imaging, we were able to successfully distinguish distinct microenvironments, based on tumor type, which influenced immune cell subpopulations and cytokine release. Although each tumor was marked by functionally distinct pathways of immune cell activation and inflammation, they also shared commonalities that ultimately resulted in granzyme B release and tumor killing. CONCLUSIONS: These results suggest that discrete tumor immune microenvironments can be identified in both responsive and non-responsive tumors and offers strategic targets for intervention to overcome checkpoint inhibitor resistance.