Holden T. Maecker, PhD, Research Professor, Microbiology & Immunology, Stanford University
Our lab is interested in finding blood-based biomarkers that could be predictive of immunotherapy success, for different tumor types and different immunotherapy settings. We use CyTOF mass cytometry, with in vitro polyclonal stimulation, to read out a broad range of cell type frequencies and functions. In two trials involving CTLA-4-based immunotherapy of advanced melanoma, we found CD8+ T cells with a central memory (CM) phenotype, or that are capable of producing IL-2, to be increased in responders compared to non-responders. Because CM T cells are capable of greater proliferation and functional differentiation than more effector-like T cells, their abundance may be one indicator of the ability to mount an effective anti-tumor T cell response. However, in anti-PD-1 treated melanoma patients, we did not find a significant correlation of CM CD8+ T cell frequencies and response. Instead, we found activated NK cells to be increased in responders vs. non-responders, suggesting a possible unexpected mechanism of action of PD-1 through NK cells. We anticipate that, while CM CD8+ T cells may be important for anti-tumor T cell immunity generally, and for CTLA-4 based therapy in particular, there will be other blood-based predictive biomarkers unique to certain immunotherapy agents.