Integrated single-cell and spatial transcriptomics reveal divergent immunological and stromal programs in peritoneal versus ovarian endometriosis

BACKGROUND: Endometriosis encompasses heterogeneous lesions, primarily peritoneal and ovarian subtypes, with distinct immunological and fibrotic features. The subtype-specific roles of the immune microenvironment in pathogenesis remain unclear. Integrated analysis combining high-resolution cellular atlases with spatial context is needed to decipher these mechanisms. METHODS: We performed an integrative analysis of published single-cell RNA sequencing (n = 81,676 cells) from control endometrium (Ctrl), eutopic endometrium (EuE), peritoneal endometriosis (EcP), and ovarian endometriosis (EcO), and spatial transcriptomic (n = 60 segments) data from Ecp and match EuE. Analytical pipelines include cellular atlas construction, differential expression, spatial co-expression, and cell-cell communication inference. RESULTS: Our single-cell atlas revealed fundamentally distinct cellular ecosystems. EcP was enriched for CCL19 + perivascular cells (6.4-fold vs. control) and immune-active niches, whereas EcO exhibited pronounced smooth muscle differentiation and NNMT upregulation. Spatial analysis compartmentalized these programs, with the CCL19-CCR7 chemokine axis localized to stromal niches in EcP strongly correlated (ρ = 0.872). In contrast, EcO showed enhanced smooth muscle gene expression (MYH11, log₂FC = 3.87). Cell-cell communication networks diverged, with EcP dominated by chemokine and TGF-β signaling, and EcO dominated by smooth muscle and PDGF pathways. Therapeutic target analysis revealed subtype-specific patterns: immune checkpoints (CTLA4 and PDCD1) were upregulated in EcP, whereas MMP9 was dramatically downregulated in EcO. CONCLUSION: EcP and EcO exhibit fundamentally divergent programs. We propose to conceptualize these as an ‘immune-hot’, chemokine-driven inflammatory niche (EcP) versus an ‘immune-cold’, fibromuscular survival niche (EcO). These findings underscore that lesion-specific immune-stromal crosstalk may dictates pathogenesis and suggest a potential paradigm shift toward immunologically-informed, subtype-specific precision therapeutics, which warrant further experimental validation.