Project 3: Predictors of sensitivity to immunotherapy and targeted treatments based on real world evidence

PROJECT 3 ABSTRACT Predictors of sensitivity to immunotherapy and targeted treatments based on real world evidence Project Leaders: Phillipe Bedard (Princess Margaret); Nikolaus Schultz, (MSK) Approvals of precision medicines directed against defined genomic targets and immune checkpoint inhibitors (ICIs) for genomic signatures of immunotherapy benefit have transformed the outcomes of many patients whose

tumors harbor these alterations. However, there are unanswered questions about targeting uncommon driver mutations, as well as the pan-cancer clinical actionability and optimal definition(s) of approved tumor agnostic genomic biomarkers, such as tumor mutation burden (TMB) for ICI therapy. While each clinical scenario where

these questions remain may be individually rare, hampering the feasibility of prospective clinical trials to address specific hypotheses, they collectively impact a substantial proportion of patients that might benefit from precision medicine approaches. The AACR Project GENIE clinical-genomic database is ideally suited for curation of

clinical outcomes of relevant patients with genomic alterations, providing essential additional evidence for prospective real-world clinical decisions in these scenarios. Our specific aims are to: 1) Characterize the clinical outcomes of patients treated with immune checkpoint inhibitor (ICI) therapy and their association with molecular biomarkers; 2) Evaluate the clinical outcomes of

patients with specific driver mutations treated with targeted therapies; and 3) Leverage observational GENIE data to improve outcome prediction in the OncoKB precision oncology knowledge base. We will investigate associations between TMB and clinical-pathological characteristics, genomic covariates, and treatment

outcomes with ICI therapy including progression free survival evaluated using PRISSMM methodology and overall survival with ICI treatment. We will analyze genomic and transcriptomic data from a subset of samples associated with outliers responses to ICI treatment. We will curate per patient efficacy outcomes for patients with

rare oncogenic driver mutations, including ERBB2 mutations and BRAF class II/III mutations, treated with matched off-label targeted therapies to investigate tumor type specificity, variant sensitivity, and the impact of co-mutations on treatment response. We will create an evidence-based framework within the OncoKB precision

medicine knowledge base to include observational data from GENIE and an outcome prediction tool for ICI treatment with clinical and genomic covariates. The ultimate goal of this Project is to use observational data from GENIE to develop a better understanding of predictors of sensitivity to immunotherapy and targeted treatments

to inform clinical decision-making.