Gene Expression Classifier Could Predict Anti-TNF Nonresponse in RA Patients

September 29, 2019
Patrick Campbell

A study examining a gene expression classifier from Scipher Medicine called PrismRA found the system could be used to identify at least half of the non-responders with at least 90% accuracy.

Results of a new study suggest gene expression profiles derived from whole blood could predict nonresponse to anti-tumor necrosis factor therapy(anti-TNF) in patients with rheumatoid arthritis.

The study, which was presented at the Clinical Congress of Rheumatology (CCR) West 2019 annual meeting in San Diego, CA, found a machine learning and network-based computation approach identified at least half of the non-responders with at least 90% accuracy.

Using predictive biomarkers identified through the Autoimmune Bio-marker Collaborative Network microarray dataset, investigators attempted to validate the effectiveness of Scipher Medicine’s cross-platform, cross-cohort gene expression classifier PrismRA. The dataset used in the study consisted of 75 biologic-naive rheumatoid arthritis patients.

Computational models used in the study were generated from the American College of Rheumatology ACR50 criteria. Investigators validated the results of their analyses using RNA sequencing data obtained from the Consortium of Rheumatology Researchers of North America.

Upon analysis, the blinded validation of the gene expression classifier yield a negative predictive value of 90% or greater and a true negative rate of 50% or greater. The results indicate PrismRA had the ability to identify at least half of the non-responders with at least 90% accuracy and could nearly double the response rate of patients receiving anti-TNF therapy.

Based on the results of the study investigators suggested gene expression profiles derived from whole blood could be used to reproducibly predict primary nonresponse to anti-TNF therapy among patients with rheumatoid arthritis.

“Clinical implementation of such a test could provide RA patients access to alternative targeted therapies as first line therapies, which may result in faster achievement of meaningful clinical change for more patients,” investigators wrote.

This study, “PrismRA™: A Cross-Platform, Cross-Cohort Gene Expression Classifier to Predict Non-Response to Anti-Tumor Necrosis Factor Therapy in Rheumatoid Arthritis Patients,” was presented at CCR West.


x