A Signature For Tumor Neoantigen-Reactive T-Cells In Fresh Human Lung Cancers Allows Rapid Cloning Of Their Receptors.

Abstract The correlations between tumor mutational load and higher response rates to immune-checkpoint inhibitor (ICI) therapies revealed the importance of targeting tumor neoantigens in cancer immunotherapy. Although ICI therapy has been revolutionary, many patients still do not respond to these treatments. One reason for this may be the absence of an adequate anti-tumor T-cell repertoire. A potential solution is the adoptive transfer of T-cells targeting neoantigens. Despite having some common shared neoantigens such as KRAS and p53, the majority of patients have highly individualized neoantigens and personalized T-cell responses. One method for identifying neoantigen-reactive T-cells from a patient's tumor-infiltrating lymphocytes (TIL) involves: (1) identifying non-synonymous mutations via whole-exome sequencing (WES), (2) using autologous dendritic cells (DC) to display these mutations in the form of electroporated minigenes (as RNA) or pulsed synthetic peptides (to create a tumor cell surrogate) and (3) co-culturing the TIL with these DCs to detect neoantigen-reactive T-cells through the upregulation of T-cell activation markers (e.g., 4-1BB). Each of these steps is labor-intensive, costly, and time-consuming. To address this, we sought to identify neoantigen-reactive TCRs directly from fresh tumors by developing a signature based on the cell surface protein and transcriptomic phenotype of cells proven to be mutation reactive. Knowing the V(D)J sequences for T-cells known to be neoantigen reactive (by the above screening), we performed 10x Genomics single cell RNA-Seq with CITE-Seq analysis on TIL isolated from frozen samples of those fresh tumors. Using the known CDR3 sequences, we compared neoantigen-reactive T-cells (six CD8 reactivities and two CD4 reactivities) versus other T-cells. This analysis led to identifying a set of cell surface proteins and genes that are specifically over- or under-expressed (including but not limited to CD62L-, CD45RA-, IL7Rlow, CD39+, CD27+, CD74+, TIGIT+, CXCL13+, LAYN+, HMOX1+, BATF+) on neoantigen-reactive T-cells. We tested whether CD39+ and CXCL13+ signature in combination with the TCR clonotype frequency could prospectively identify other neoantigen-reactive TCRs. This proved to be the case and applied not only to CD8 TIL but strongly to CD4 reactivities. In all cases, we could identify neoantigen-reactive TCRs, and in one case, 8 out of 10 candidate receptors identified by phenotype signature proved to be neoantigen-reactive. Not only were they neoantigen specific, but new, previously unidentified neoantigens were found when their specificities were analyzed. This method can open the path to treating patients with custom-made TCRs in a timely manner and expand the repertoire of tumor-specific neoantigens and the diversity of the T-cell response discoverable in human tumors. Citation Format: Kenichi Hanada, Chihao Zhao, Raul Gil-Hoyos, Jared Gartner, Christopher Chow-Parmer, Frank Lowery, Sri Krishna, Samuel Chatmon, Prickett D. Todd, Scott Kivitz, Maria Parkhurst, Michelle Langhan, Thomas Shelton, Zulmarie Franco, Robert Somerville, John Wunderlich, David Danforth, Zachary Rae, Kelly Michael, Nathan Wong, Paul Robbins, Steven Rosenberg, James Yang. A signature for tumor neoantigen-reactive T-cells in fresh human lung cancers allows rapid cloning of their receptors [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 1509.