Pan-cancer extracellular vesicle (EV) liquid biopsy signatures as predictors of response to immune checkpoint inhibitors
Aim/goals:
Our objective is to recruit from the Atlantic Node Biobanks a retrospective and prospective cohort of 30 patients for each of the following tumor site: head and neck cancer (HNSCC), triple negative breast cancer (TNBC), endometrial cancer (EC), kidney cancer (KC) and upper gastro intestinal (UGI) that have or will receive immune checkpoint therapy. From patient plasma samples, we isolate the extracellular vesicles (EVs) and analyze by RNA sequencing, cytokine arrays, flow cytometry and MS proteomics for differential expression or levels of relevant biomarkers. Artificial intelligence algorithms will be performed on this multi-omic data set and findings will be validated on prospective samples from patients undergoing ICI therapy.
Summary:
Immune checkpoint inhibitors (ICIs) are changing the landscape of cancer care. Therapies that target PD- 1/PD-L1 are increasingly used as first line therapy for many cancer types. Unfortunately, even in the best- case scenario, response to ICI therapy rarely surpasses 50% of patients, and for some cancer types, little or no benefit is observed. A better understanding of factors that influence response to ICI treatment is required. EVs released from both tumour and immune cells in the tumour microenvironment have been found to play an important role in immune evasion. We have previously identified EV components that are likely involved this process including targetable immune suppressive factors. Our goal is to identify similar patterns in other cancers and validate these findings using immune response models. We believe this will lead to novel treatment combinations to overcome ICI treatment failure in cancer patients.
From the ICI treated patient cohort, EVs, collected from peripheral blood samples, will be queried using multiple omics approaches to compare responders to non-responders. This project will employ artificial intelligence to mine datasets of patient ICI response signatures using both tumour whole genome sequencing (WGS) and transcriptomic profiling as well as EV transcriptomic, protein and cytokine profiling from patient plasma. Identified inhibitory factors will be validated and targeted strategies to overcome inhibition will be developed. Thereafter, we will use our established humanized mouse model system to test whether the phenotype of tumor cells can be predicted by circulating EVs and ICI. The cohorts will be sufficiently powered to facilitate analysis of response profiles to ICI therapy.
