LSC Abstract – Multi-level differential network analysis of COPD exacerbations

Rationale: Patients with chronic obstructive pulmonary disease (COPD) often suffer episodes of exacerbation (ECOPD) that impact negatively the course of their disease. We hypothesized that the comparison of multi-level (i.e., clinical, biological, imaging and microbiological) correlation networks during ECOPD and at clinical recovery can yield useful pathogenic information. Aims: To describe the weight and interactions among the different multi-level network components of ECOPD. Methods: In a multicenter study in Spain, we prospectively measured 67 clinical, biological, imaging and microbiological variables in 76 COPD patients when hospitalized because of ECOPD and during clinical stability, and used multi-level correlation networks to compare both conditions. Results: Main results showed that, compared with clinical stability, during ECOPD there is disruption of co-regulated lung function and systemic inflammation network modules. High circulating neutrophils, serum glucose and CRP (and low levels of blood lymphocytes and eosinophils) are the panel of biomarkers that best describe the ECOPD condition. Differences between sputum inflammation levels at exacerbation and recovery defined three subtypes of ECOPD: (1) without pulmonary inflammation; (2) with higher pulmonary inflammation and infectious markers during ECOPD; and, surprisingly, (3) with increased inflammation and infectious markers at clinical stability. Conclusions: This is the first study that uses multi-level, differential network analysis to investigate the complex relationship between different core components of ECOPD. Results identify network disruption during ECOPD, a panel of biomarkers associated with ECOPD and three different ECOPD subtypes.