Integrative Analysis Reveals a Molecular Stratification of Systemic Autoimmune Diseases

Abstract

Objective Clinical heterogeneity, a hallmark of systemic autoimmune diseases (SADs) impedes early diagnosis and effective treatment, issues that may be addressed if patients could be grouped into a molecular defined stratification. Methods With the aim of reclassifying SADs independently of the clinical diagnoses, unsupervised clustering of integrated whole blood transcriptome and methylome cross‐sectional data of 955 patients with 7 SADs and 267 healthy controls was undertaken. In addition, an inception cohort was prospectively followed for 6 and 14 months to validate the results and analyze if cluster assignment changed or not with time. Results Four clusters were identified and validated. Three were pathological representing ‘inflammatory’, ‘lymphoid’, and ‘interferon’ patterns each including all diagnoses and defined by genetic, clinical, serological, and cellular features. A fourth cluster with no specific molecular pattern associated with low activity, and accumulated also healthy controls. A longitudinal and independent inception cohort showed a relapse‐remission pattern, where patients remained in their pathological cluster, moving only to the healthy one, thus showing that with time, the molecular clusters remain stable and that single pathogenic molecular signatures characterize each individual patient. Conclusions Patients with SADs can be jointly stratified into three stable disease clusters with specific molecular patterns differentiating different molecular disease mechanisms. These results have important implications for future clinical trials and the study of therapy non‐responsiveness marking a paradigm shift in our view of SADs.