转化型滤泡性淋巴瘤肿瘤微环境的单细胞空间多组学表征

Histological examination of follicular lymphoma (FL) biopsies remains the cornerstone for diagnosis and staging of FL. Single cell sequencing approaches, while transcriptomically rich, require tissue dissociation and therefore lose the native spatial context that may underpin FL progression and transformation to a high-grade lymphoma, typically diffuse large B cell lymphoma (DLBCL). By single cell spatial profiling, we aimed to characterize the spatiotemporal heterogeneity and interplay between malignant B cells and the tumor microenvironment (TME) during transformation from FL to DLBCL histology. Single cell spatial transcriptomics (Nanostring CosMx) and proteomics (Cell DIVE) were performed on FFPE tissue cores of 14 tFL patients with paired FL (tFL-FL) and DLBCL (tFL-DLBCL) timepoints, and 13 FL patients without evidence of progression or transformation (non-tFL; >6y of follow-up). Spatial omics confirmed malignant B cell dominated follicular structures in non-tFL and tFL-FL while absent in tFL-DLBCL, consistent with pathologic classification. Follicles were then delineated (defining intra-follicular, peri-follicular and extra-follicular regions) and enriched proliferation and exhaustion features were observed in intra-follicular regions compared to extra-follicular counterparts. Moreover, differential spatial interaction patterns were observed, amongst which Galectin-9 crosstalk components were identified as being differentially expressed and exhaustion-related pathways were more prevalent in intra-follicular and peri-follicular regions in non-tFL and tFL-FL timepoints. Macrophage phenotypes also markedly shifted across spatial regions with progression: intrafollicular FOLR2⁺ and IL4I1⁺ macrophages present in non-tFL progressively diminished, while CD163⁺, SPP1⁺, and NLRP3⁺ macrophages became increasingly enriched in extra-follicular regions in tFL-DLBCL. Cell neighborhood analysis showed differential abundance among non-tFL, tFL-FL, and tFL-DLBCL timepoints. For B-cell-predominant cell neighborhoods (CNs), the germinal center (GC) score was higher in non-tFL than tFL-FL, whereas tFL-FL and tFL-DLBCL shared similar GC score levels. Within CNs of similar cell type compositions, T cells were more exhausted and macrophages were more polarized towards anti-inflammatory phenotypes in the transformed state. Subsequent single cell proteomics validated the differential follicular Galectin-9 signaling and macrophage-related gene expression patterns among spatial regions. In summary, we delineated the spatial landscape of both indolent and transformed FL. We identified a spatiotemporal shift in TME composition, spatial interactions, and cell neighborhoods. This study provides novel insights into the spatial lymphoma architecture associated with FL transformation, contributing to a refined disease evolution model.