Abstract
Cancer immunotherapy is now an established treatment option for many cancers. Cancer immunotherapy boosts host antitumor immunity to provide long-term benefit; however, only a small fraction of the treated patients show a durable clinical response. The tumor microenvironment ecosystem, with its complex mixture of non-malignant and malignant cells, is a major contributor in regulating the response to checkpoint blockade and development of resistance. Ongoing efforts to characterize the tumor microenvironment to stratify patients for immunotherapy and find biomarkers of response often use methods that are limited by the availability of adequate tumor tissue from needle biopsy material and loss of tissue viability during sample processing that precludes the use of single-cell sequencing platforms. Therefore, genomic methods that use deconvolution to assess the relative proportion of different cell types and their phenotypes in the tumor microenvironment are desirable for clinical use. To this end, we have developed OncoPeptTUME, a novel computational approach that utilizes a proprietary minimal gene expression signature to assign immune scores for eight broad categories of immune cells present in the tumor microenvironment. To validate the approach, we used 9,640 TCGA gene expression datasets from 33 different tumors, defined their immune cell content, organized samples into clusters based on their immune cell content, and identified the molecular differences that predict survival in samples belonging to different clusters. We further performed a deeper analysis of samples enriched in infiltrating CD8+ T-cells to identify T-cell phenotype that correlated with a long-term survival benefit. A small set of genes associated with functional T-cell phenotype was used on a dataset of melanoma samples to show that higher expression of the genes discriminated responders from the nonresponders to nivolumab treatment. In conclusion, our analysis demonstrates that OncoPeptTUME is a powerful immunogenomic tool to predict patient prognosis, stratify patients who will benefit from cancer immunotherapy and identify pathways and novel biomarkers of long-term benefit from the use of cancer immunotherapy drugs.
Citation Format: Xiaoshan Shi, Malini Manoharan, Nitin Mandloi, Sushri Priyadharshini, Laxman Iyer, Rohit Gupta, Papia Chakraborty, Amitabha Chaudhuri, Ravi Gupta. OncoPeptTUME: A novel computational approach analyzes the tumor microenvironment to predict response to checkpoint inhibitors [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr A112.