Background/Purpose: Calcinosis cutis affects up to 20% of adults with dermatomyositis (DM), causing significant morbidity including recurrent infections, incapacitating pain, and functional impairment. Current management focuses on surgical excision despite the risk of recurrence. To better understand and treat this devastating complication, we must first better understand its pathogenesis, including gene expression profiles and relevant cell types.

Methods: We performed Visium 10X spatial transcriptomics on four calcinosis samples from adult DM patients meeting 2017 ACR/EULAR criteria (positive for anti-MDA5, PM-Scl, Ro, and Jo-1) and three healthy adult control skin biopsies. All DM samples contained active calcinosis defined as new or evolving lesions. Data preprocessing was conducted using cellranger, with downstream analysis conducted in R using the Seurat v5.0.0 package. Pseudo-bulk analysis was conducted using the DESeq2 v. 1.44.0 package, and all other differential expression was conducted using the default FindMarkers command. Spatial annotations were conducted in Loupe Browser 8 by one observer (CP).

Results: Pseudo-bulk analysis (Fig. 1) identified a distinct inflammatory signature with elevated IL6 RNA expression and upregulation of macrophage markers (CD68, MSR1) and chemokine CCL3. We observed increased transcription of RNA encoding immunoglobulins (IGKC, IGHG1, IGHA1, JCHAIN). Extracellular matrix dysregulation was evidenced by decreased keratin expression, elevated MMPs 1/9/13, and elevated cartilage/bone-specific collagens. Osteopontin (SPP1), which resorbs pathological calcification, was robustly elevated in lesional samples. UMAP analysis revealed minimal overlap between DM and healthy control samples, prompting us to adopt a spatial-first approach. In the DM samples, spatial transcriptomic spots were categorized based on their distance from the calcinosis lesion into four groups: < 50 µm (C), 50–100 µm (P), 100–150 µm (2P), and 150–200 µm (3P). Differential expression analysis across these groups showed that SPP1 expression was highest near the calcinosis, while professional antigen-presenting cell (APC) markers, immunoglobulin genes, and complement genes were most enriched 100–150 µm from the lesion. Additionally, a complex interplay of matrix remodeling genes, including MMPs and their regulators (TIMPs), was observed at varying distances from the calcinosis (Fig. 2).

Conclusion: This study provides the first detailed spatial molecular map of calcinosis in DM, suggesting a complex pathogenic cascade involving macrophage-driven inflammation, enhanced local antibody production, and increased osteopontin expression. Further validation is planned to refine our interpretation and explore cell types and signals driving disease pathogenesis.

Heatmap showing gene expression fold changes between DM-associated calcinosis and healthy skin samples from pseudobulk analysis of Visium spatial transcriptomic data. Genes selected are the top most significantly differentially expressed (all adjusted p-value < 0.001).

Violin plots showing gene expression patterns across tissue regions at varying distances from calcinosis deposits in DM samples. C = calcinosis, P = perilesional ( < 50um from calcinosis), 2P = secondary perilesional (50-100um from calcinosis), 3P = tertiary perilesional (100-150um from calcinosis). * = adjusted p-value < 0.05, ** = adjusted p-value < 0.001

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/spatial-transcriptomic-analysis-of-calcinosis-cutis-in-dermatomyositis-uncovers-disease-associated-pathways-involving-il-6-tissue-remodeling-and-osteopontin/