It is now well accepted that patients with inflamed tumors are more likely to benefit from immunotherapies targeting immune checkpoints such as CTLA4, PD1 and PD-L1.Several factors determine a tumor's immune status and identifying patients who may benefit from immunotherapy and determining which immune therapy to administer depends largely on the patient's tumor immunophenotype. We have developed an approach for immunoprofiling using a focused gene set (115 genes) and scoring system that assesses cell specific types and immune functions. This approach was developed in a preclinical cancer models and is being evaluated and tested in human cancers. We previously extracted data for the gene set and employed unsupervised network clustering to differentiate and classify cancer samples of patients from The Cancer Genome Atlas (TCGA) into immunologically distinct clusters. Here continue or examination and overview clinical features associated with individual clusters. A total of 8477 patients from the TCGA research network were profiled which included 28 solid cancer types. A total of 10 immune clusters (IC 1-10) were identified with a median cluster size of 971 patients with sizes ranging from 178 to 1570. All cancer types were represented in the ICs with mostly normal distribution except for IC 6 which was composed of 40% breast cancer patients and IC10 of which 1/5 of the cases were head-neck squamous cell carcinomas. ICs 2, 5 and 8 were associated with the poorest survival with median times at 73, 51 and 65 months, respectively. Conversely, patients in IC 9 experienced the longest overall survival, median time 148 months. The aggregate inflammation score, which measures overall tumor inflammation was positively correlated to overall survival (P=0.0109). IC 9 was correlated with the highest survival times and was positively associated with primary signatures related antitumor immunity such as high IFN gamma, cytotoxicity, and CD8 T cell and NK cell signatures. IC 8 was characterized with poor survival times and was associated with signatures related with immunosuppressive cells including granulocytic myeloid derived suppressor cells, M2 polarized macrophages and TH17 T cells, and an adenosine high signature. In addition, cases from this cluster also had low CTL signatures. Our findings show that gene expression data from a focused gene and our scoring algorithm could be used to differentiate and classify patients into various groups of immunologically distinct tumors types.

Citation Format: Marco A. De Velasco, Yurie Kura, Kazuko Sakai, Kazutoshi Fujita, Masahiro Nozawa, Kazuhiro Yoshimura, Kazuto Nishio, Hirotsugu Uemura. Gene panel-based immune profiling of human cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 622.

©2021 American Association for Cancer Research.
2021
American Association for Cancer Research.