Published in:
Open Access
01-12-2017 | Research
Multiple biomarkers predict disease severity, progression and mortality in COPD
Authors:
Rachel L. Zemans, Sean Jacobson, Jason Keene, Katerina Kechris, Bruce E. Miller, Ruth Tal-Singer, Russell P. Bowler
Published in:
Respiratory Research
|
Issue 1/2017
Login to get access
Abstract
Background
Chronic obstructive pulmonary disease (COPD) is a heterogeneous disease characterized by multiple subtypes and variable disease progression. Blood biomarkers have been variably associated with subtype, severity, and disease progression. Just as combined clinical variables are more highly predictive of outcomes than individual clinical variables, we hypothesized that multiple biomarkers may be more informative than individual biomarkers to predict subtypes, disease severity, disease progression, and mortality.
Methods
Fibrinogen, C-Reactive Protein (CRP), surfactant protein D (SP-D), soluble Receptor for Advanced Glycation Endproducts (sRAGE), and Club Cell Secretory Protein (CC16) were measured in the plasma of 1465 subjects from the COPDGene cohort and 2746 subjects from the ECLIPSE cohort. Regression analysis was performed to determine whether these biomarkers, individually or in combination, were predictive of subtypes, disease severity, disease progression, or mortality, after adjustment for clinical covariates.
Results
In COPDGene, the best combinations of biomarkers were: CC16, sRAGE, fibrinogen, CRP, and SP-D for airflow limitation (p < 10−4), SP-D, CRP, sRAGE and fibrinogen for emphysema (p < 10−3), CC16, fibrinogen, and sRAGE for decline in FEV1 (p < 0.05) and progression of emphysema (p < 10−3), and all five biomarkers together for mortality (p < 0.05). All associations except mortality were validated in ECLIPSE. The combination of SP-D, CRP, and fibrinogen was the best model for mortality in ECLIPSE (p < 0.05), and this combination was also significant in COPDGene.
Conclusion
This comprehensive analysis of two large cohorts revealed that combinations of biomarkers improve predictive value compared with clinical variables and individual biomarkers for relevant cross-sectional and longitudinal COPD outcomes.