Spatial transcriptomic profiling of selected facial cutaneous lupus erythematosus and rosacea with overlapping features.

BACKGROUNDCutaneous lupus erythematosus (cLE) and rosacea are generally distinct clinicopathological entities, but selected facial inflammatory presentations may show overlapping features. Molecular biomarkers may provide adjunctive information in such equivocal cases.

OBJECTIVETo identify candidate molecular biomarkers distinguishing cLE from rosacea using spatial profiling techniques.

METHODSFFPE skin biopsy samples from cLE and rosacea patients were analyzed using GeoMx™ Digital Spatial Profiling platform. Differentially expressed genes were identified, and pathway analysis was conducted. Candidate biomarkers were validated using immunofluorescence. An exploratory logistic regression model was developed using selected biomarkers and evaluated with leave-one-out cross-validation.

RESULTSISG15 and OASL were upregulated in cLE, whereas TMEM9, MOCS3, SPNS2, and TMCO1 were elevated in rosacea at both transcriptomic and protein levels. Logistic regression selected OASL, ISG15, TMEM9, and MOCS3 as a four-marker exploratory panel, which showed preliminary within-cohort discriminatory value in leave-one-out cross-validation but was not externally validated. In two diagnostically ambiguous cases, biomarker staining profiles were concordant with subsequent clinical course.

CONCLUSIONSpatial transcriptomic profiling identified candidate biomarkers that may serve as molecular adjuncts for distinguishing rosacea from selected facial cLE cases, predominantly ACLE/SCLE in this cohort. These findings are exploratory and require validation in larger independent cohorts, including broader cLE subtypes, before clinical application.

Abstract from PubMed