CELL INTELLIGENCE · 肿瘤细胞治疗研究

Tertiary lymphoid structures (TLS) have evolved from descriptive histological findings into candidate biomarkers and targetable immune niches in solid tumors. In intrahepatic cholangiocarcinoma (ICC), however, similar lymphoid architectures may have different biological implications depending on their relationship to tumor nests, immune circuitry, and suppressive stroma. We define a "TLS topotype" as the multidimensional state of an individual TLS-centered region, including the TLS core, its perilesional neighborhood, and the nearest tumor-stroma interface. This state integrates anatomic topology and tumor accessibility, structural maturity, putative effector circuitry, and suppressive context; treatment-induced plasticity is considered separately through longitudinal assessment. Because heterogeneous TLS may coexist within one tumor, the whole tumor is represented by a "topotype landscape" based on compartment-specific TLS density and the distribution of TLS-level states rather than a single label. Direct human ICC evidence supports compartment-resolved associations between TLS distribution and outcome and identifies suppressive B-cell, Treg, myeloid, and stromal programs. These observations establish spatial TLS organization as prognostically relevant, but do not validate the framework or demonstrate treatment-specific prediction. By contrast, TLS-resident TCF7-positive stem-like CD8 T-cell maintenance, coordinated B-cell-dendritic-cell-T-cell circuitry, and ICI-specific predictive value remain unproven in ICC and should be considered hypothesis-generating. The framework therefore distinguishes candidate circuit-competent, tumor-accessible TLS states from stromally excluded or suppressive-neighborhood states while allowing mixed profiles. Multiplex pathology and spatial transcriptomics may define these states, whereas radiomics, liquid biopsy, and pretreatment biopsy provide complementary but incomplete information. Biopsy can characterize only TLS states captured in sampled tissue, and radiomics currently estimates whole-lesion TLS probability rather than functional topology. Marker expression or spatial colocalization alone does not establish functional productivity, which requires evidence of tumor-antigen presentation or specificity, local lymphocyte renewal or clonal expansion, functional perturbation, or a spatially linked treatment response. By separating TLS-level state, tumor-level landscape, and patient-level trial stratum, this review provides a framework for biomarker development, therapeutic reprogramming, prospective validation, and longitudinal monitoring in ICC. No TLS topotype has yet been validated to predict differential benefit from immune checkpoint inhibition in ICC.