This proposal addresses PQ2: How do variations in immune function caused by comorbidities or observed among different populations affect response to cancer therapy? While the role of the immune system in controlling and eradicating tumors has been known for decades, it is only in the past few years that immunotherapy has emerged as a clinically viable way to target cancer. Little is known about how the immune response differs between populations defined by gender and ethnicity, either in terms of levels of immune infiltration or the intrinsic functionality of specific immune cell types. We are developing resources and methods to dissect the impact of levels of infiltrating leukocytes on cancer outcomes. PRECOG is a large compendium of gene expression datasets comprising nearly 40,000 patient samples across 39 distinct cancer histologies, for which overall survival information (and other outcomes) is available. Previously we used PRECOG to assess the influence of immune cell levels on overall survival, identifying commonalities and differences across cancer. Many of these associations were therapy-independent, emphasizing how the immune system is a critical component of patient response even in the context of standard chemotherapy. Here we will extend PRECOG to incorporate information on patient gender and ethnicity. No published studies have systematically analyzed similarities and differences in the impact of immune cell types on clinical outcomes in these groups. Here we identify how infiltrating immune levels vary between tumors from different populations, and how they differentially affect responses to specific therapies as well as overall survival. This will generate testable hypotheses regarding variations in immune infiltration and function between populations. These will also be associated with response to specific therapies, and with overall survival. Using novel deconvolution approaches we will further dissect immune function in relation to survival and therapy response. This prognostic map will illuminate similarities and differences across the landscape of cancer populations and will be a powerful new resource for the cancer immunologic community.
We aim to identify gender- and ethnicity-specific prognostic factors influencing survival outcomes for cancer patients from genomic data. Building on the PRECOG resource, we will develop a map of prognostic associations that is available to the wider community. This will also provide a ?road map? for identifying potential molecular mechanisms driving differential therapy response in different sub-populations of patients.
