384 Features of neoantigen-reactive T cells with <i>ex vivo</i> expansion capacity
Bibliographic record
Abstract
Background A high number of neoantigen-reactive tumor-infiltrating T lymphocytes (TIL) for adoptive transfer immunotherapy is a key mediator of clinical responses in metastatic melanoma and gastrointestinal cancers. 1 2 However, the baseline molecular features defining neoantigen-reactive TIL with ex vivo proliferative capacity remain poorly defined, which in turn limits the development of novel manufacturing methods and patient-selection criteria.Methods We analyzed 21 tumors from patients with colorectal, lung, ovarian, and melanoma cancers. Primary expansion of TIL from tumors was obtained with IL-2, followed by a standard bulk rapid expansion protocol (bulkREP) with irradiated allogeneic PBMC feeders, OKT3 and IL-2, and in parallel, a REP of TIL sorted based on their reactivity to predicted neoantigen peptides pulsed on autologous APC (neoREP). Single-cell CITEseq, RNA and TCR-seq ( n = 13) were performed at baseline in TIL from fresh tumors combined with flow cytometry and bulk TCR-seq, and ex vivo expansion was tracked with bulk TCR sequencing after primary expansion and REP.Results Flow cytometry and bulk TCR-seq revealed substantial clonotype attrition during REP, especially among CD8 + T cells, regardless of enrichment for neoantigen-reactivity. From the baseline single-cell analysis of 41,966 TIL, the starting frequency of predicted neoantigen-reactive TIL inferred from the most validated published gene signatures 3 was not predictive of neoantigen reactivity after primary expansion, bulkREP, or neoREP. However, neoantigen-reactive and ex vivo proliferative TIL (neoTILprolif) arose from specific cell clusters. In CD4+ T cells, follicular-like T cells expressing CXCL13 represented the main source, while in CD8+ T cells, clusters of already proliferative T cells and of GZMK-expressing effector memory cells sourced most cells, contrasting with a low input from tissue-resident memory T cells. Genes most highly expressed in neoTILprolif had little overlap with gene signatures that only predicted neoantigen reactivity. By differential protein expression analysis, the integrin α2 (CD49b) and PD-1 significantly distinguished both CD4+ and CD8+ neoTILprolif from neoantigen-reactive TIL that contracted in cultures, as well as from proliferative, naïve-like, bystanders.Conclusions Neoantigen-reactive TIL with ex vivo proliferative potential constitute a minority of T cells in tumors and have a unique transcriptomic profile and cell surface protein expression to guide antigen-agnostic manufacturing of T cells more likely to mediate clinical response.Acknowledgements This work was supported by preclinical research funding from Turnstone Biologics.References Kristensen NP, Heeke C, Tvingsholm SA, et al. Neoantigen-reactive CD8+ T cells affect clinical outcome of adoptive cell therapy with tumor-infiltrating lymphocytes in melanoma. J Clin Investig 2022;132:e150535.Lowery FJ, Goff SL, Gasmi B, et al. Neoantigen-specific tumor-infiltrating lymphocytes in gastrointestinal cancers: a phase 2 trial. Nat Med 2025;1–10.Lowery FJ, Krishna S, Yossef R, et al. Molecular signatures of antitumor neoantigen-reactive T cells from metastatic human cancers. Science 2022;375:877–84.Ethics Approval This study was approved by the Centre hospitalier de l’Université de Montréal Institutional Review Board, approval number 21.299.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".