On-line breath analysis with secondary electrospray ionization discriminates between COPD patients with and without frequent exacerbations
M. T. Gaugg, Y. Nussbaumer-Ochsner, L. Bregy, A. Engler, N. Stebler, T. Bruderer, P. M-L Sinues, R. Zenobi, M. Kohler
Introduction: On-line breath analysis is a painless, non-invasive method to get insights into the metabolism of patients. Using real-time mass spectrometry, we investigated the profiles of volatile and semi-volatile organic compounds (VOC, sVOC) in exhaled breath of patients suffering from chronic obstructive pulmonary disease (COPD) with and without frequent exacerbations. The aim of the study was to discriminate between the two groups as well as to get insights into underlying metabolic differences and to better understand the pathogenesis of exacerbations.
Methods: COPD patients with frequent and non-frequent exacerbations were recruited and matched in terms of gender, age and smoking history. Breath analysis was performed in real-time, using secondary electrospray ionization mass spectrometry (SESI-MS). Extensive data analysis was then applied to identify metabolic patterns that differ between the two groups and to assess their classification performance.
Results: We successfully identified metabolic patterns in exhaled breath, which discriminate COPD patients with and without frequent exacerbations. Using a leave-one-out cross validation (LOOCV) we obtained a ROC-AUC of 0.89. In total, we observed 380 significant (q < 0.05) features, many of which formed correlating chemical families. Amongst others, a series of nitro-aromatic compounds was increased in frequent exacerbators. We propose that these findings correlate with elevated nitric oxide production in response to the pulmonary inflammation. Other significant groups were different short and medium-chained fatty acid derivatives.
Conclusions: On-line breath analysis is capable of discriminating COPD patients with and without frequent exacerbations with good accuracy. It is a painless, non-invasive method that has the potential to provide information on biochemical processes in contrast to conventional lung function tests. We propose that our findings correlate with the increased oxidative stress caused by elevated nitric oxide production in response to the pulmonary inflammation.