Spatially resolved immune niches in thyroid cancer: from hot-cold-excluded ecosystems to precision immunotherapy.

Although the overall prognosis of most thyroid cancers is relatively good, the benefits of immunotherapy in advanced, dedifferentiated, and some special subtypes still show significant heterogeneity. The existing evaluation frameworks based on PD-L1, tumor mutational burden, or conventional transcriptomic signals are insufficient to explain the complex and variable immune response patterns among different patients and within the same tumor. In recent years, single-cell sequencing, spatial transcriptomics, and related spatial multi-omics studies have shown that the immune microenvironment of thyroid cancer is not a homogeneous background but is composed of multiple local ecological niches with clear spatial organizational characteristics. These ecological niches have significant differences in cell composition, functional state, and interaction mode. The current evidence suggests that the regions rich in B cells and tertiary lymphoid structures in papillary thyroid carcinoma are often associated with relatively indolent clinical behaviors; undifferentiated thyroid carcinoma more frequently presents as an inhibitory spatial pattern characterized by macrophages, cancer-associated fibroblasts, and immune exclusion boundaries; and the neural-immune crosstalk in medullary thyroid carcinoma further indicates that some "cold" immune phenotypes may be actively shaped by neuroendocrine signals. From the perspective of spatial immune niches, this article re-examines the biological basis and translational significance of hot, cold, and excluded immune patterns in thyroid cancer, and discusses their potential implications for immune classification, biopsy strategies, and the optimization of precise immunotherapy.