Subsets of Patients With Lupus Identified by Gene Expression Profiles Exhibit Differential Clinical Responsiveness to Baricitinib.
Prathyusha Bachali, Amrie Grammer, Peter E Lipsky
✦ AI-curated · Sources linked
In 30 seconds
This phase 2 randomised controlled trial analysis examined whether baseline gene-expression profiles could identify patients with systemic lupus erythematosus (SLE) who respond to baricitinib. Whole-blood transcriptomes from 272 participants were grouped into eight molecular endotypes, and clinical response was assessed after 24 weeks of placebo or baricitinib. Significant responses versus placebo were confined to endotypes D and G, with reported effect sizes of 41.54% and 31.89%, respectively. Clustering by clinical features did not identify responsive subsets, whereas interferon and Treg-cell signatures were important predictors of response.
Key findings
- Eight molecular endotypes were identified, with progressively increasing immune disturbances across groups.
- Significant responses to baricitinib versus placebo occurred only in endotypes D and G (P = 0.004 and 0.048; effect sizes 41.54% and 31.89%).
- Endotype G showed dose-dependent suppression of interferon, JAK1/JAK2/TYK2, immunoproteasome, and inflammatory pathways from week 2 through week 24.
- Interferon and Treg-cell signatures were identified as major predictors of response.
Why it matters
The findings suggest that molecular heterogeneity in SLE may help explain why treatment effects are not uniform across trial populations. Transcriptomic endotyping identified response-associated subsets that clinical-feature clustering did not detect, supporting further study of molecular markers as a way to refine treatment-response assessment in SLE.
What to keep in mind
The analysis was based on a phase 2 trial and its findings do not establish that transcriptomic endotyping predicts response in other populations or trials.