Document resource
Background Adoptive transfer of autologous T cells from tumor fragments cultured in IL-2 has been successful in treating melanomas and select epithelial cancers. 1 2 However, tumor-reactive T cells are a small percentage of all TIL while most are bystanders.3 Identification of neoantigen-reactive TIL for treatment involves time and resource-intensive testing of TIL cultures against patient-specific candidate neoantigens, resulting in selection of TIL at the population rather than cell level.4 Use of cell surface markers to isolate tumor-relevant TIL will circumvent the need for personalized sequencing and screening and generate a more polyclonal TIL product.Methods In this study, we cultured multiple fragments from resected tumors, together in GREX100 flasks with high dose IL-2 (n=9) obtaining a median of 4.4e8TIL. The large cell yield allowed the study of cell surface marker expression for sorting tumor-reactive TIL by FACS. To identify candidate markers, we performed single-cell RNA-, CITE- and TCR sequencing on TIL from a human rectal cancer lung metastasis. We cloned and expressed high frequency CD8TCRs in donor lymphocytes and screened them for tumor reactivity. Twelve TCRs showed functional reactivity to tumor neoantigens and 3 were reactive to tumor-irrelevant CMV/EBV/Flu peptides (indicating true bystanders). Projecting relevant TCRs onto the transcriptomic map indicated that cells with tumor-reactive TCRs clustered separately from bystanders. Antibody expression analysis showed CD103 as the top marker for reactive cells and CD31 for bystanders. Subsequently we evaluated these markers to isolate tumor-reactive CD8+TIL.Results Among the 9 patients studied, the percentage of CD103+CD31- cells ranged from 4.6%-44% among CD8+TIL. TIL sorted based on CD103/CD31 expression were expanded using the rapid expansion protocol. Subsequent screens against candidate neoantigens demonstrated enriched reactivity among CD103+CD31-CD8+TIL in 7/9 patients compared to bulk CD8+T cells. Mean functional reactivity (upregulation of 4-1BB following on-target stimulus) among the CD103+CD31-CD8+TIL was 20.7% and median fold enrichment was 5.7X over bulk CD8+TIL. Single-cell TCR sequencing analysis performed on the reactive T cells enabled identification and validation of 1-3 unique neoantigens per patient, in some cases with multiple clonotypes against the same neoantigen. Further, sorting CD103+CD31-CD8+T cells enriched for cells with stem-like phenotype 5 indicating their enhanced anti-tumor potential.Conclusions This study has strong implications for the identification of reactive CD8+TIL from bulk cultures of tumor fragments without requiring patient-specific testing. Pending successful validation of scalability of this approach under GMP conditions, it could substantially decrease time-to-treatment (~5 weeks) making continued exploration of TIL therapy viable for more epithelial cancer patients.References Parkhurst M, et al. Adoptive transfer of personalized neoantigen-reactive TCR-transduced T cells in metastatic colorectal cancer: phase 2 trial interim results. Nature Medicine 2024;30(9):2586-2595.Lowery FJ, et al. Neoantigen-specific tumor-infiltrating lymphocytes in gastrointestinal cancers: a phase 2 trial. Nature Medicine 2025.Lowery FJ, et al. Molecular signatures of antitumor neoantigen-reactive T cells from metastatic human cancers. Science 2022;375(6583):877-884.Parkhurst MR, et al. Unique neoantigens arise from somatic mutations in patients with gastrointestinal cancers. Cancer Discov 2019;9(8):1022-1035.Krishna S, et al. Stem-like CD8 T cells mediate response of adoptive cell immunotherapy against human cancer. Science 2020;370(6522):1328-1334.Ethics Approval This study was approved by National Institutes of Health (NIH) Internal Review Board (IRB); approval number NCT00068003.