Integrative genomic biomarkers for predicting response to immunotherapy in melanoma patients Currently, the most effective treatment for melanoma is immune checkpoint blockade with single agent or combination immunotherapy (nivolumab and ipilimumab). However, about 70% of patients do not respond to these therapies, and many experience immune-related toxicities, such as colitis, pneumonitis, hepatitis and endocrine disorders. A number of studies have shown that clinical response is dependent on a variety of factors, including the composition of immune cells in the tumor microenvironment, PDL1 expression, antigen processing and presentation, tumor expression of interferon-responsive genes and possibly somatic mutations, among others. Nevertheless, the genomic features that best discriminate immunotherapy responders from non-responders remain unclear, and practical methods for measuring such predictive biomarkers in clinical settings are lacking. The main goal of this proposal is to develop an integrated framework for profiling genes and cell types that predict response to immunotherapy in patients with melanoma. We are proposing to perform deep genomic profiling of DNA, RNA, and immune composition in a large previously banked cohort of tumor biopsies, plasma samples, and peripheral blood leukocytes obtained from melanoma patients before, during, and after checkpoint blockade. Our major distinguishing feature is the cohort size, and the use of our recently described analytical methods for (1) profiling circulating tumor DNA with high sensitivity and specificity and (2) determining immune composition in tumors and blood samples using only gene expression data. The former will allow tumor genetic heterogeneity to be analyzed directly from the circulation, enabling the detection of genetic changes linked to immunotherapy without the need for invasive biopsies. The latter enables the enumeration of immune composition in complex tissues without requiring cell dissociation, surface markers, or antibodies.
In Aims 1 and 2, we will interrogate our cohort for novel genomic signatures of immune cell composition, gene expression, and tumor genetic aberrations that predict response to immunotherapy at baseline and during treatment.
In Aim 3, we will integrate these features with known molecular markers and will evaluate their relative contributions to patient outcomes. Our findings will be the basis for an optimized laboratory test that predicts response to nivolumab and ipilimumab, and will serve as a model for other diseases. Clinical impact: The results of our studies will assist both research and clinical laboratories in performing a novel diagnostic assay that incorporates robust genomic markers predictive of patient outcomes. Thus far, the tests offered by major commercial entities are restricted to recurrent mutations in known melanoma driver genes for which targeted therapy is available, as described in NCI-MATCH (the Actionable Mutations and Matching Drugs by the NCI-Molecular Analysis for Therapy Choice Trial). None of these genes have been implicated in response to melanoma immunotherapies. Our proposal directly addresses this challenge.
The research of this proposal will construct immunotherapy-related biomarkers (nivolumab and ipilimumab) for profiling melanoma patients' response to the drugs. It will contribute to the development of an integrative molecular test for the diagnosis of patients under immunotherapy. All the molecular data will be deposited in SRA/dbGAP and cBioPortal so they can be incorporated in computers designated to analyze data from cancer patients receiving immunotherapy.
