Claim: RA biologic-response biomarkers should be stratified by drug class, not pooled

Claim: rheumatoid arthritis biologic-response biomarkers are drug-class-specific enough that anti-IL-6R and anti-TNF should be modeled separately rather than pooled into one universal baseline panel.
Evidence: In the MONARCH biomarker analysis, baseline and week-2 serum predictors differed between sarilumab (anti-IL-6R) and adalimumab (anti-TNF). The anti-IL-6R signal tracked more with innate immune activation and synovial inflammation, while the anti-TNF signal tracked more with T-cell and neutrophil biology. A separate adalimumab whole-blood transcriptomics study identified 84 response-associated genes and highlighted MZB1 as a novel biomarker, reinforcing that TNFi response has its own molecular structure.
Inference: If the clinical goal is pretreatment drug selection, a pooled "biologic response" panel will likely blur mechanism and underperform class-specific gates.
Limitations: Both sources are biomarker-discovery studies with limited absolute cross-validated performance. This supports stratified model development, not standalone clinical deployment.
References:
- Agueusop I, et al. Deciphering differential biomarkers for anti-interleukin-6 receptor and anti-tumour necrosis factor-α treatment response in rheumatoid arthritis by multiomics analysis. DOI: 10.1136/rmdopen-2025-005556.
- Yap CF, et al. Identifying Predictive Biomarkers of Response in Patients With Rheumatoid Arthritis Treated With Adalimumab Using Machine Learning Analysis of Whole-Blood Transcriptomics Data. DOI: 10.1002/art.43255.