Spatial transcriptomic analysis of chronic lymphocytic leukemia lymph nodes reveals tumor microenvironment evolution during disease progression

Abstract Introduction . Chronic lymphocytic leukemia (CLL) cells bidirectionally interacts with the tumor microenvironment (TME). Leukemic cells establish within lymph nodes (LNs) a pro-tumoral niche that drives disease progression and therapy resistance. A deeper understanding of dynamic CLL-TME modifications is essential to identify novel markers and actionable cell-to-cell interactions for more specific targeted therapies. We applied Digital Spatial Profiling (DSP) to characterize tumor and immune cells, including T lymphocytes and macrophages, in CLL-infiltrated LNs, keeping distinct microanatomical areas present within proliferation centers (PCs) and from those pertinent to the surrounding non-PC. Methods. Seventeen FFPE CLL-LNs were grouped into 3 clinical subsets: A (n=6) indolent CLL; B (n=4) CLL requiring treatment within 6 months after LN biopsy; C (n=7) relapse/progressive CLL. Spatial transcriptomics was performed using the GeoMx® Digital Spatial Profiler. Region of interest (ROIs) were identified and selected based on multiplexed immunofluorescence with anti-CD3 anti-CD68 and MUM1 for the identification of PCs T cells, macrophages, and PC neoplastic B cells, respectively. ROIs inside and outside PCs were selected for each cell type. Normalized transcriptomic matrices were generated for each cell segment, and Gene Ontology (GO) analysis was performed on pre-ranked differentially expressed genes (DEGs; *p*-value < 0.05, |log₂ fold change| > 0). Pathways were selected based on an FDR Q-value threshold (Q < 0.05), NOM p-value (< 0.05) and ranked by Normalized Enrichment Score (NES). Results DSP was performed on a total of 295 ROIs, comprising 88 CD3+ ROIs, 106 CD68+ ROIs, 101 PC neoplastic B cells ROIs and distributed within or outside PCs. First, transcriptomic profile of pooled ROIs within and outside PC were compared for each cell type. As expected CLL ROIs within PC showed upregulation of MUM1, MYC, and PCNA genes, thus providing evidence of ongoing leukemic cell proliferation. CD3⁺ T lymphocytes inside PCs exhibited enrichment in pathways associated with proliferation and pro-inflammation, and upregulation of STAT1, NF-kB1, and CCL19 genes compared to non-PCs ROIs. As for macrophages, they showed enrichment in genes associated with oxidative stress (SOD1, TRX2), chromatin remodelling and inflammatory pathways, and showed upregulation of the checkpoint molecule LAG3. Our subsequent analysis focused on comparing cell-specific transcriptomic pattern associated with the three distinct clinical subgroups. In group A, CD3+ T lymphocytes retain an immunological, activated, and homeostatic-like transcriptional profile, with enrichment of pathways related to positive immune system process, cell activation and signalling, along with upregulation of key genes including TNF, IL17D, STAT5B, NF-kB1, IFNA16. In group B, CD3+ T lymphocytes displayed upregulation of genes like ENO1, AKT2, PTPN1 associated with metabolism and cellular activation, vesicle trafficking, mitochondrial activity, and chromatin remodelling. In group C, a defective T cell status became apparent (enriched pathways: apoptosis, downregulation of cell function, motility, negative transcription, with main genes including CR1, MYB, IL3RA, IL12A, TNFRSF21). CD68+ macrophages showed an increased functional state in group A, with enriched pathways related to immunity and motility. However, in more aggressive phases of the disease (groups B and C) they showed enrichment in cell activation, chromatin remodelling, vesicle trafficking and mitochondrial activity. In group A, tumor cells were enriched in B cell activation, antigen-receptor mediated signalling pathway, and immune response regulating signalling pathways. In group B, we observed pathways associated with cell cycle, chromatin remodelling, vesicle trafficking and organelles, and motility while, in the group C, vesicle trafficking and organelles, along with inflammation-related pathways, were prominent. Conclusions Our findings demonstrate that within CLL-LNs, both TME and leukemic cells activate distinct transcriptional programs based on their location inside or outside the anatomically defined MUM1+ PCs, which function as proliferative hubs. Moreover, our transcriptomic analysis reveals that, as CLL clinically progresses, tumor and TME co-evolve, shifting from a more physiological condition of immunological activity, toward an increasingly dysfunctional and immunologically altered status.