Intratumoral microbiota and immune predictors of response to immunotherapy in lung cancer

ABSTRACT Although immunotherapy, especially immune checkpoint inhibitors (ICIs), has emerged as a powerful cancer treatment, less than half of advanced Non-Small Cell Lung Cancer (NSCLC) patients have responded. There is a crucial need for biomarkers to enable better prediction before ICI therapy and to overcome ICI resistance.

Among the contributing factors to ICI therapy resistance, the immunosuppressive nature of the tumor microenvironment (TME) is one of the most challenging. Despite the important roles of both intratumoral microbiota and tumor-infiltrating immune cells, there are huge gaps in linking specific tumor-residing microbiota

changes with immune cell subpopulations in NSCLCs treated with ICIs. We hypothesize that an integrative prediction model that incorporates intratumoral microbiota and immune predictors has the potential to substantially distinguish ICI responders from non-responders, and improve the selection of patients that are most likely to benefit from ICI therapy. We propose to capitalize

on existing prospectively collected pre-ICI tumors or biopsies in ICI-treated NSCLCs from three cohorts (n = 500): Baylor College of Medicine cohort, Boston Lung Cancer Survival cohort, and Moffitt Cancer Center cohort. Patients will be classified as ICI responders (defined as complete/partial response and stable disease) and non-

responders (defined as progression), using the modified Response Evaluation Criteria in Solid Tumors (mRECIST) criteria. To identify new predictors that distinguish responders from non-responders before ICI therapy, we propose the following three Specific Aims: 1) To identify intratumoral microbiome profiles predictive

of ICI response, using Whole-Metagenome Sequencing (WMS); 2) To characterize spatial immune signatures predictive of ICI response, using single-cell Imaging Mass Cytometry (IMC); and 3) To develop integrated predictive models of ICI response incorporating clinical, microbial, and immunological data using deep learning

approaches. Integration of tumor microbiome and its interactions with immune infiltrate within TME is a relatively new field of investigation and challenging endeavor, which has not been reported in NSCLCs. This proposed study will build upon archived formalin-fixed, paraffin-embedded specimen repositories (pre-treatment tumor resection and core

biopsies), mature clinical treatment response and survival data from three independent cohorts, experience with the proposed experimental microbiome and spatial immune profiling (WMS and IMC) approaches, institutional core facilities, and the multidisciplinary research team, including cancer epidemiologists, immunologists,

microbiologists, oncologists and bioinformatics experts.