A multi‐omic single cell sequencing approach to develop a CD8 T cell specific gene signature for anti‐PD1 response in solid tumors

Immune checkpoint blockade (ICB) has led to durable clinical responses in multiple cancer types. However, biomarkers that identify which patients are most likely to respond to ICB are not well defined. Many putative biomarkers developed from a small number of samples often fail to maintain their predictive status in larger validation cohorts. We show across multiple human malignancies and syngeneic murine tumor models that neither pretreatment T cell receptor (TCR) clonality nor changes in clonality after ICB correlate with response. Dissection of tumor infiltrating lymphocytes pre‐ and post‐ICB by paired single‐cell RNA sequencing and single‐cell TCR sequencing reveals conserved and distinct transcriptomic features in expanded TCR clonotypes between anti‐PD1 responder and nonresponder murine tumor models. Overall, our results indicate a productive anti‐tumor response is agnostic of TCR clonal expansion. Further, we used single‐cell transcriptomics to develop a CD8+ T cell specific gene signature for a productive anti‐tumor response and show the response signature to be associated with overall survival (OS) on nivolumab monotherapy in CheckMate‐067, a phase 3 clinical trial in metastatic melanoma. These results highlight the value of leveraging single‐cell assays to dissect heterogeneous tumor and immune subsets and define cell‐type specific transcriptomic biomarkers of ICB response. What's new? Immune checkpoint inhibitors have achieved dramatic results in a subset of patients, but it's still difficult to predict which patients will respond. Here, the authors developed a gene signature associated with response to anti‐PD‐1 therapy. They compared changes in immune cell populations between syngeneic mouse tumor models that responded to anti‐PD1 and those that did not. Using single‐cell transcriptomics, they developed a CD8+ T cell‐specific gene signature that correlates with anti‐PD1 response. To demonstrate clinical relevance, they retrospectively analyzed human orthologs in the dataset from the large CheckMate‐067 trial and found that the anti‐PD1‐responsive expression profile correlated with better survival.