Abstract 2121: Saliva TCR repertoire as a tool for head and neck cancer immunophenotype monitoring

Abstract

Abstract Head and neck squamous cell carcinoma (HNSCC) is a global health problem. Successful immunotherapy in HNSCC is currently based on the ability to block the interaction of PD1 and PDL1 and abolish the inhibition of CD8+ T cells, thus enhancing the antitumor activity. T cells recognize peptide antigens in the context of MHC molecules through the clonally distributed T Cell receptor (TCR), which offers the means to detect and track specific T cells. We characterized TCR signatures in unique HNSCC paired samples (including L = Lymphocyte, S = Saliva, T = Tumor) to profile TCR repertoire diversity and clonality in different compartments. A total of 48 HNSCC samples corresponding to 14 patients (2 assayed 2X) were obtained from Johns Hopkins University. DNA samples were submitted for TCR sequencing (TCR-seq) via enrichment of the human TCRB locus using Human-TCRB-PD4bx. CDR3 sequences were downloaded from Adaptive ImmuneAnalyzer. For each sequence, V, D and J regions were extracted and translation to amino acid done on the CDR3 region. 8 paired (T, L, S) samples were submitted for shallow sequencing survey depth and 8 for deep assay depth. In order to better understand the TCR repertoire, we assessed the number of unique immune receptor clonotypes, their relative abundance, repertoire overlap, clonotype tracking and sequence length distribution. As expected, the deep assay revealed more clones than the survey. Using the TCR-seq deep assay, 9 out of 10 S samples harbored at least 10% of the number of clonotypes found in their paired T. As expected, L harbored a greater number and more diverse clones than T and S samples and showed partial overlap with T and S samples. Comparing the overlap between T, L and S based on a shared CDR3.aa repertoire: by shallow survey, 3/5 S samples had at least 1 overlap with the primary T. Using the deep assay, 9 out of 10 patients (90%) showed overlap between the S and paired T. We found top public clonotypes occupying the larger part in HPV negative samples, and surprisingly no known CDR3.aa sequences annotated to HPV in HPV positive samples but annotated to other common viruses (CMV, EBV). Our study demonstrates a robust TCR repertoire in S that corresponds at least partially to the clones observed in T and circulating L. The presence of HPV did not yet yield CDR3.aa specific sequences. Future single-cell transcriptomes may identify signatures of CD8+ and CD4+ neoantigen-reactive tumor-infiltrating L in the paired S samples in addition to the viral or tumor-associated antigens present in bulk assays. Our data need to be further validated in larger well characterized cohorts that include S from cancer-free subjects. These early results support the use of saliva as a potential surrogate for diagnostic immune-profiling of tumors, for early detection and follow-up. We further envision creating a personalized TCR repertoire for individual patients from an initial tumor sample biopsy to monitor their tumor immune dynamics using saliva. Citation Format: Dieila Giomo De Lima, Mariana Brait, Laura Palmieri, Fernando T. Zamuner, Esther Broner, Kellie N. Smith, Timothy Westlake, Or Malca, Sol Efroni, Ido Sloma, David Sidransky. Saliva TCR repertoire as a tool for head and neck cancer immunophenotype monitoring [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2121.