Alleviated T cell exhaustion and SLC1A3-mediated stroma-remodelling dictate chemoimmunotherapy efficacy in oesophageal squamous cell carcinoma

Background Combining chemotherapy with anti-programmed cell death protein-1 (PD-1) improves clinical outcomes in oesophageal squamous cell carcinoma (ESCC), yet the underlying synergistic mechanism remains obscured. Moreover, 30–50% of patients still derive no therapeutic benefit from the combination strategy, highlighting the need to decipher and overcome resistance. Objective We sought to investigate the mechanisms by which chemotherapy augments the responses to immune checkpoint blockade and elucidate the factors contributing to persistent resistance in non-responding patients. Design We designed a systematic investigation involving longitudinal sampling of ESCC tissues both from patients treated with chemotherapy plus anti-PD-1 and anti-PD-1 monotherapy. The tumour microenvironment (TME) was then comprehensively characterised using single-cell transcriptomics, T cell receptor repertoire analysis, multiplex immunohistochemistry and murine models. Results We demonstrated that combination therapy exerted superior antitumour efficacy by mitigating immune checkpoint engagements (TIGIT-NECTIN2 and NECTIN1-CD96) between epithelial-stress tumour cells and CD8+ T cells, thereby preventing T cells from exhaustion and boosting vitality. In non-responders, we identified a subset of tumour cells with high SLC1A3 expression, which localised at the tumour boundary and interacted with COL1A1+ myofibroblastic cancer-associated fibroblasts, inducing an extracellular matrix-enriched TME that hindered the infiltration of CD8+ T cells. Inhibiting SLC1A3 significantly enhanced the efficacy of chemotherapy plus anti-PD-1, underscoring its potential as a therapeutic target. Conclusion This study elucidates the synergistic mechanisms and identifies key resistance pathways underlying chemo-immunotherapy combinations in patients with ESCC, providing a scientific basis for refining future combination therapeutic regimens. Data are available in a public, open access repository. Data are available upon reasonable request. The curated high-quality scRNA-seq data, scTCR-seq of ESCC patient samples and mouse scRNA-seq data were deposited at in the Genome Sequence Archive at the National Genomics Data Center with accession numbers: PRJCA037418. The raw sequencing data of validated ESCC scRNA-seq cohort and bulk RNA-seq of Li et al 2023 ESCC cohort could be obtained in the Genome Sequence Archive at the National Genomics Data Center with accession numbers: PRJCA012636[20][1]. Any additional information required to reanalyse the data reported in this paper is available from the lead contact upon reasonable request. [1]: #ref-20