717 Enhanced molecular signatures in cutaneous lupus erythematosus patients support distinct pathogenic pathways in African American patients

Abstract

Cutaneous lupus erythematosus (CLE) is an autoimmune disease with clinical sequelae such as skin damage that disproportionately affects African Americans (AAs). Recently, a novel approach in gene expression analyses was devised that groups genes into transcriptional modules (i.e. apoptosis, protein synthesis, inflammation) to identify relevant gene signatures. This technique helped identify a unique interferon signature in systemic lupus patients. Thus, we applied this modular analysis approach to identify molecular signatures unique to AA CLE patients. We conducted RNA sequencing of whole blood transcriptomes from 66 CLE patients (52% (N=34) AAs) and subsequently performed modular analyses. Modules were associated with patient subgroups distinguished by demographic, clinical, and laboratory features. An unsupervised cluster analysis identified eight distinct clusters of CLE patients. Statistical analyses comparing module scores of these clusters were performed using Kruskal-Wallis tests with Bonferroni’s correction. We observed that two groups with mostly AA CLE patients (n=12 (35.3% of AAs)) had a predominant T cell signature (M4.1, M4.15 (both p<0.0001)). Seven inflammation module scores (M3.2, M4.2, M4.6, M4.13, M5.1, M5.7, M7.1 (all p<0.0001)) were markedly elevated in two clusters of mainly non-AA patients (N=16 (51.6% of non-AAs)). Additionally, neutrophil (M5.15) and cell death (M6.13, M6.6 (all p≤0.0001)) molecular signatures were up-regulated in these two patient clusters suggesting that neutrophils and cell death processes could be driving the inflammatory response in non-AA CLE patients. Thus, our data suggests that unique cell populations and processes may be driving disease pathophysiology in AA and non-AA CLE patients, potentially affecting their disease course and treatment selection.