Abstract 2787: Landscape of tumor-infiltrating B and T cell repertoire in pancreatic cancer

  1. Pineda, Silvia 2
  2. Yu, Katharine 3
  3. Malats, Nuria 1
  4. Sirota, Marina 3
  1. 1 3CNIO, Madrid, Spain.
  2. 2 1CNIO, MAdrid, Spain;
  3. 3 2UCSF, San Francisco, CA;
Revista:
Cancer Research

ISSN: 0008-5472 1538-7445

Año de publicación: 2021

Volumen: 81

Número: 13_Supplement

Páginas: 2787-2787

Tipo: Artículo

DOI: 10.1158/1538-7445.AM2021-2787 GOOGLE SCHOLAR lock_openAcceso abierto editor

Otras publicaciones en: Cancer Research

Resumen

B and T cells are a key component of adaptive immunity with diverse functions including antibody production, antigen presentation, and cellular cytotoxicity. They are required to recognize cells expressing foreign or mutated proteins, including neoantigens from cancer cells. Infiltrating B and T cells have been observed in several tumor tissues, but their role remains unknown. Pancreatic cancer (PC) is a dreadful disease and, despite its relatively low population incidence, it is the deadliest cancer worldwide with 7% 5-year survival rate. The majority of known risk factors point to a chronic inflammatory process suggesting that different forms of inflammation play decisive roles in tumor development resulting in immunological infiltration in the tumor microenvironment. It is thought that highly mutated tumors are more likely to harbor neo-antigens, which make them targets for adaptive immunity. Furthermore, PC is resistant to immunotherapy and the reasons are unknown. Therefore, characterizing the tumor immune infiltration repertoire can help to elucidate the pro- or anti- tumor responses in PC. We used MiXCR tool to map the read sequences from RNA sequencing data to their respective B-cell receptors (BCR) and T-cell receptors (TCR) clonotypes. We used 144 PC samples from TCGA and 180 pancreatic tissue from healthy donors from GTEx. We found that the diversity (Shannon Entropy) of BCR and TCR clonotypes were significantly higher (p-value < 2.2*10-16) in tumoral than in normal tissue suggesting that the immune infiltration might be regulated through tumor-specific mutations. In addition, we defined a network for each sample considering the BCR-TCR clone definition (same V and J segments, same CDR3 length, and 90%-95% nucleotide identity between CDR3s) and found that tumor samples were more clonally expanded than normal samples for BCR. We did not find clonal expansion for TCR which was negatively associated with low neoantigens load, suggesting that T cells are present in PC but inactive, therefore an effective immune response cannot be generated from an immune suppressive tumor microenvironment. On the other hand, higher levels of BCR are related with higher levels of tobacco and also with a better overall survival. These results point to differences in the immunological infiltration pathways in PC. Being able to determine which PC have more BCR and TCR infiltration and whether this is due to different forms of inflammation or other genetic or environmental exposure factors could improve our understanding of pancreas carcinogenesis towards a better patient stratification and prevention, as well as to facilitate development of potential strategies for treatment.