Panel of Biomarkers Can Distinguish Stable, Exacerbation States in COPD

Artificial neural network model can predict exacerbations at median of seven days before clinical diagnosis



TUESDAY, Nov. 26, 2024 (HealthDay News) — A panel of biomarkers can distinguish between stable and exacerbation states in chronic obstructive pulmonary disease (COPD), and an artificial neural network (ANN) model can predict exacerbations before the occurrence of symptoms, according to a study published online Nov. 20 in ERJ Open Research.

Ahmed J. Yousuf, from the University of Leicester in the United Kingdom, and colleagues measured 35 biomarkers implicated in COPD pathogenesis in paired urine samples from 55 COPD patients during stable and exacerbation states. A model combining the 10 most discriminatory biomarkers was developed as a near-patient dipstick test. This biomarker panel was tested in 105 COPD patients who underwent daily home urine testing over six months. An ANN was developed based on urine biomarkers from 85 of the 105 patients and was tested for predicting exacerbation risk.

The researchers found that the 10-biomarker panel could significantly distinguish exacerbation versus stable state in the discovery and validation studies (receiver operating characteristic with an area under the curve, 0.84 and 0.81, respectively). The ANN model predicted an exacerbation within a 13-day window frame, with an area under the curve of 0.89. Exacerbations were identified at a median of seven days before clinical diagnosis.

“Our findings suggested that the identified urine biomarkers are promising in discriminating COPD exacerbations from stable COPD, can be measured at home with a lateral flow reader and mobile technology and so could be used at scale,” the authors write.

Several authors disclosed ties to Mologic, which funded the study.

Abstract/Full Text

Page 1 of 1