Abstract
Nuclear-magnetic-resonance (NMR) profiling of exhaled breath condensate (EBC) provides insights into the pathophysiology of bronchiectasis by identifying specific biomarkers. We evaluated whether NMR-based metabolomics discriminates the EBC-derived metabolic phenotypes ("metabotypes") of 41 patients with non-cystic fibrosis (nCF) bronchiectasis of various etiology [24 subjects with Primary Ciliary Dyskinesia (PCD); 17 patients with bronchiectasis not associated with PCD (nCF/nPCD)], who were compared to 17 healthy subjects (HS). NMR was used for EBC profiling, and Orthogonal Projections to Latent Structures with partial least-squares discriminant analysis (OPLS-DA) was used as a classifier. The results were validated by using the EBC from 17 PCD patients not included in the primary analysis. Different statistical models were built, which compared nCF/nPCD and HS, PCD and HS, all classes (nCF/nPCD-PCD-HS), and, finally, PCD and nCF/nPCD. In the PCD-nCF/nPCD model, four statistically significant metabolites were able to discriminate between the two groups, with only a minor reduction of the quality parameters. In particular, for nCF/nPCD, acetone/acetoin and methanol increased by 21% and 18%, respectively. In PCD patients, ethanol and lactate increased by 25% and 28%, respectively. They are all related to lung inflammation as methanol is found in the exhaled breath of lung cancer patients, acetone/acetoin produce toxic ROS that damage lung tissue in CF, and lactate is observed in acute inflammation. Interestingly, a high concentration of ethanol hampers cilia beating and can be associated with the genetic defect of PCD. Model validation with 17 PCD samples not included in the primary analysis correctly predicted all samples. Our results indicate that NMR of EBC discriminates nCF/nPCD and PCD bronchiectasis patients from HS, and patients with nCF/nPCD from those with PCD. The metabolites responsible for between-group separation identified specific metabotypes, which characterize bronchiectasis of a different etiology.