Clonotype neighbor graph analysis (CoNGA) reveals tumor-infiltrating population of EBV antigen-specific T-cells in cutaneous T-cell lymphoma

Summary: Abstract Body: Despite the morbidity associated with cutaneous T-cell lymphoma (CTCL), its pathophysiology remains incompletely understood, and a definitive antigen driver has yet to be determined. We aimed to identify reactive tumor-infiltrating lymphocytes (TILs) in CTCL tumor skin biopsies and characterize their antigen specificities. To this end, we utilized Clonotype Neighbor Graph Analysis (CoNGA), a novel computational tool for uncovering functionally related groups of T-cells, to analyze publicly-available single-cell TCR and gene expression sequencing data. From a cohort of 10 patients with Mycosis Fungoides, three were found to harbor TILs with TCRs specific for Epstein-Barr Virus (EBV) epitopes. In two of these patients, an EBV-specific clone was found to be among the 15 most expanded clones within the skin tumor sample. Gene expression analysis revealed upregulation of genes such as GZMA, GZMK, and CCL5, suggesting a cytotoxic effector function of these clones. These findings highlight a potential role for EBV antigens in previously-infected patients with CTCL, possibly contributing to enhanced lymphocyte stimulation by these epitopes within the tumor microenvironment. A broader analysis of EBV-reactive T cells in a larger cohort of CTCL patients could provide deeper insights into their role in anti-tumor immunity. Additionally, this study demonstrates the utility of CoNGA in uncovering antigen-specific T-cells within heterogeneous tumor samples. The algorithm allows for the discovery of T-cell populations beyond the malignant clone that share antigen specificity and function, providing a powerful means of understanding anti-tumor immune dynamics and identifying novel therapeutic targets. Alexander Kaminsky¹, Annemieke de Jong², Sara Suhl¹, Larisa J. Geskin² 1. Columbia University Vagelos College of Physicians and Surgeons, New York, NY, United States. 2. Dermatology, Columbia University, New York, NY, United States. Bioinformatics, Computational Biology, and Imaging