A Digital Evaluation of T-Cell Infiltration of Human Cancers and Its Use as a Prognostic and Predictive Tool

Abstract

T-cell signatures are used to evaluate T-cell infiltration of non-lymphoid tissues in experimental and clinical settings. However, none of the T-cell signatures reported so far are optimal. This paper describes the development of a new human T-cell signature comprising 15 genes that are overexpressed at least tenfold in resting and activated T cells as compared to non-lymphoid tissues and cell lines. We demonstrate that signature-H shows high sensitivity for evaluating T-cell infiltration in healthy tissues and in the 32 cancers investigated. Further, signature-H works as well as immunohistochemistry methods for assessing the prognosis of neuroblastoma. Finally, we used a multiparametric tool that includes T-cell infiltration based on the new signature to predict the response of melanoma to the anti-PD-1 antibody nivolumab and obtained positive results. These preliminary findings suggest that the new T-cell signature can be used as a prognostic tool and in tailoring treatment, in the precision medicine perspective.

Keywords: tissue-resident memory T cells, T cell infiltration of human cancer, microarray, prognostic tool, predictive tool, precision medicine