Analytical and clinical validation of an RNA sequencing-based assay for quantitative, accurate evaluation of a molecular signature response classifier in rheumatoid arthritis

Disease Module Detection

Alex Jones, Sarah Rapisardo, Lixia Zhang, Theodore Mellors, Johanna B. Withers, Zoran Gatalica, and Viatcheslav R. Akmaev


Objectives: This study reports analytical and clinical validation of a molecular signature response classifier (MSRC) that identifies rheumatoid arthritis (RA) patients who are non-responders to tumor necrosis factor-ɑ inhibitors (TNFi).

Materials and Methods: The MSRC integrates patient-specific data from 19 gene expression features, anti-cyclic citrullinated protein serostatus, sex, body mass index, and patient global assessment into a single score.

Results: The MSRC results stratified samples (N = 174) according to non-response prediction with a positive predictive value of 87.7% (95% CI: 78–94%), sensitivity of 60.2% (95% CI: 50–69%), and specificity of 77.3% (95% CI: 65–87%). The 25-point scale was subdivided into three thresholds: signal not detected (<10.6), high (≥10.6), and very high (≥18.5). The MSRC relies on sequencing of RNA extracted from blood; this assay displays high gene expression concordance between inter- and intra-assay sample (R2 > 0.977) and minimal variation in cumulative gene assignment diversity, read mapping location, or gene-body coverage. The MSRC accuracy was 95.8% (46/48) for threshold concordance (no signal, high, very high). Intra- and inter-assay precision studies demonstrated high repeatability (92.6%, 25/27) and reproducibility (100%, 35/35).

Conclusions: The MSRC is a robust assay that accurately and reproducibly detects an RA patient’s molecular signature of non-response to TNFi therapies.

View Full Paper