Models created with traditional + novel biomarkers have clinically meaningful predictive power
TUESDAY, Aug. 2, 2016 (HealthDay News) — Low abundance biomarker panels can predict lupus nephritis outcomes, according to research published online July 27 in Arthritis & Rheumatology.
Noting that the American College of Rheumatology guidelines recommend change in lupus nephritis induction therapy when no response has occurred within six months, Bethany J. Wolf, Ph.D., from the University of South Carolina in Charleston, and colleagues analyzed biomarker models to develop a tool for defining response. In 140 patients with biopsy-proven lupus nephritis who had not yet started induction therapy, the authors analyzed urine samples for a panel of urinary biomarkers. For each individual biomarker, they generated univariate receiver operating characteristic (ROC) curves, and compared them to the ROC area under the curve values from models developed using random forest algorithms.
The researchers found that there was clinically meaningful predictive power for models developed with the combined traditional and novel biomarker panels. Chemokines, cytokines, and markers of cellular damage were the most predictive of response.
“This is the first study to demonstrate the power of low-abundance biomarker panels and machine learning algorithms for predicting lupus nephritis outcomes,” the authors write. “This is a critical first step in research to develop clinically meaningful decision support tools.”
Subject materials were provided by Genentech and Bristol-Myers Squibb.
Abstract
Full Text (subscription or payment may be required)
Copyright © 2016 HealthDay. All rights reserved.
