CELL INTELLIGENCE · 肿瘤细胞治疗研究

Immune checkpoint blockade has changed the therapeutic landscape of cervical cancer, but clinical benefit remains heterogeneous and difficult to predict with single biomarkers. Cervical cancer is a particularly suitable setting for immune-cell profiling because HPV-driven antigenicity, chemoradiotherapy-induced tissue remodeling, anti-angiogenic combinations, and repeated exposure to PD-1/PD-L1 inhibitors all shape the blood-tumor immune axis. This Mini Review discusses how circulating immune cells and tumor-infiltrating immune cells can be jointly interrogated by multi-omics technologies to identify biomarkers of response and resistance. We focus on single-cell RNA sequencing, T-cell receptor/B-cell receptor (TCR/BCR) profiling, cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq), flow cytometry, spatial transcriptomics, high-plex spatial proteomics, bulk transcriptomics, cytokine profiling, and circulating tumor DNA (ctDNA)-informed longitudinal designs. Recent cervical cancer atlases suggest that response is unlikely to be explained by total immune infiltration alone, but instead reflects the spatial organization and functional states of cytotoxic T cells, NK cells, antigen-presenting cells, tertiary lymphoid structure (TLS) programs, regulatory T cells, myeloid cells, and cancer-associated fibroblasts (CAFs). We propose a blood-to-tumor framework in which dynamic peripheral phenotypes may monitor systemic immune competence, while tissue multi-omics may identify local immune exclusion, terminal exhaustion, and myeloid suppression. Prospective paired sampling and interpretable, externally validated AI-assisted computational integration are needed before these biomarkers can guide treatment selection.