Abstract Rationale Sarcoidosis is an inflammatory disease of unknown cause characterized by granulomatous immune cell aggregates across multiple organs. While granulomas exhibit a broadly conserved histological architecture, a spatially resolved understanding of their cellular organization and organ-specific features remains incomplete. Here, we systematically map the high resolution single-cell architecture of sarcoidosis granulomas across seven organs to identify shared and tissue-specific cellular states and transcriptional programs, with the goal of advancing precision mechanistic insights into granuloma biology. Methods We analyzed formalin-fixed paraffin-embedded sarcoidosis tissue samples containing granulomas from three centers (Hannover n = 82, Aachen n = 39, Paris n = 7; total n = 128) spanning lung, skin, lymph node, liver, kidney, heart, and bone marrow. All patients had a confirmed sarcoidosis diagnosis according to current guidelines. Samples were profiled using the 10x Visium HD spatial transcriptomics platform. Nuclei segmentation was performed using StarDist, followed by conversion of spatial bins into single-cell transcriptomes. Cell-by-gene matrices were integrated at the organ level. To resolve spatial transcriptional gradients, we caclulated a ring-to-core distance metric quantifying expression dynamics across granuloma architecture. Multinucleated giant cells were manually annotated to characterize their transcriptional signatures. Results Spatial single-cell analysis revealed pronounced intra-granulomatous myeloid heterogeneity with multiple macrophage and monocyte states that organized into distinct zonal layers within the granuloma. SPP1+ monocyte derived macrophages represented a conserved macrophage phenotype present not only in pulmonary granulomas (3.94% ± 12.69% cells per sample) but across all organs, suggesting a cross-organ sarcoidosis granuloma program. Surrounding stromal zones contained distinct fibroblast subsets, including inflammatory fibroblasts at the granuloma border, indicating a coordinated myeloid-stromal niche. Multinucleated giant cells exhibited a specialized macrophage program enriched for lysosomal lipid-handling and mitochondrial stress-adaptation pathways. Spatially resolved ligand-receptor analysis revealed pronounced immuno-fibrotic crosstalk between myeloid cells and adjacent inflammatory fibroblast populations surrounding the granuloma. Finally, we identified a sequence of transcriptional programs aligned with the ring-to-core axis. Validation on the protein level using multiplex immunofluorescence staining is currently underway. Conclusions This multi-organ spatial single-cell atlas reveals a conserved cellular blueprint of sarcoidosis granulomas with a structured myeloid differentiation continuum aligned along a ring-to-core axis and an inflammatory fibroblast boundary niche. These findings establish a foundational reference for granuloma biology, highlight shared versus tissue-specific programs across organs, and may serve as basis for future targeted therapeutic strategies in sarcoidosis. This abstract is funded by: Ann Theodore Foundation
Ruwisch et al. (Fri,) studied this question.