Bronchial transcriptomics identifies distinct endotypes of severe COPD

<bold>Introduction:</bold> COPD is a heterogeneous condition, characterized by varying combination and/or severity of emphysema, airways disease and mucus hypersecretion. No pharmacological treatments are available that can halt disease progression. More insight in the underlying molecular mechanisms driving the different pathological features of COPD is needed, in particular for severe COPD, the most susceptible group. <bold>Aim:</bold> To identify different COPD endotypes and elucidate which genes and biological pathways are involved. <bold>Methods:</bold> Bronchial brush transcriptomics data of 123 severe and 23 mild-moderate COPD patients included in the SHERLOCK study was used. We performed unsupervised hierarchical clustering on all COPD patients based on 219 severe COPD-specific genes. Differentially expressed genes grouped in gene modules were identified between the clusters, followed by Gene Ontology enrichment analysis. <bold>Results:</bold> Clustering separated the mild-moderate and severe COPD patients and identified 3 distinct severe COPD-associated gene expression(S-CAGE) clusters, labeled S-CAGE 1, S-CAGE 2 and S-CAGE 3. In comparision,S-CAGE 1 exhibited dysregulation of genes related to extracellular matrix organization and immune response pathways. S-CAGE 2 was characterized by lower expression of cell cycle genes, while S-CAGE 3 showed higher expression of cell cycle and DNA repair genes. <bold>Discussion:</bold> Our findings emphasize the molecular heterogeneity within severe COPD. Importantly, they unveil the presence of different subgroups based on molecular differences, i.e.endotypes, within a group of patients with similar clinical phenotype. These endotypes reflect different biological pathway activities, providing new possible targets for precision medicine.