A CADSET WP4 transcriptomic analysis of Asthma and COPD overlap

<b>Rationale:</b> Clinical features of asthma and COPD are present in some adults with airways disease (Asthma-COPD overlap, ACO). To understand the biological basis of ACO, we contrasted transcriptomic data from blood and sputum from ACO and non-ACO patients derived from an asthma (U-BIOPRED) and a COPD (ECLIPSE) cohort. <b>Methods:</b> We used weighted gene correlation network analysis (WGCNA) to interrogate transcriptomic data against 59 clinical and immune features. U-BIOPRED defined ACO as subjects with post-bronchodilator FEV1&lt;80% predicted and postbronchodilator FEV1/FVC ratio&lt;0.7 whereas ECLIPSE defined ACO as subjects with postbronchodilator change in FEV1&gt;12% or 200ml. <b>Results:</b> In U-BIOPRED we found 31 and 75 differentially expressed genes (DEGS) between ACO and nonACO blood and sputum, respectively; WGCNA modules associated with ACO. In contrast, in ECLIPSE there were no DEGs between ACO and -nonACO subjects and no WGCNA modules were associated with ACO in blood or sputum. Yet, the mediumpurple3 module was associated with blood eosinophilia and eosinophilia-related genes such as IL5RA and Siglec8. In ECLIPSE. <b>Conclusions:</b> There are significant transcriptomic differences between U-BIOPRED-ACO and -nonACO subjects but not between ECLIPSE-ACO and -nonACO subjects. According to transcriptomic data asthma ACO may exist but COPD ACO is unlikely, although there is an interesting eosinophil signal in the latter.