Effects of treatment and withdrawal with inhaled beclomethasone/formoterol on electronic nose and NMR metabolomic breathprints in patients with COPD

We performed a prospective pilot study in 14 COPD ex-smokers (females/males 2/12, age 73.6±1.8 years, mean±SEM, FEV1 61.8±6.9% pred) to investigate the effect of inhaled corticosteroid (ICS) treatment and interruption on volatile/nonvolatile compound breathprints that correlate with respiratory inflammation (Fens N et al. Eur Respir J 2011;38:1301-9). At all 4 visits, exhaled breath analysis with two different technology electronic noses (e-noses) (Cyranose 320: 32 sensors; Libranose: 8 sensors), exhaled breath condensate (EBC) collection for NMR spectroscopy, spirometry, sputum induction, and fraction of exhaled nitric oxide (FENO) measurement was performed. Patients on fluticasone propionate/salmeterol fixed dose combination (DPI) (500/50 μg bid) for at least 8 weeks (visit 1) were consecutively treated with inhaled beclomethasone/formoterol microparticle formulation (pMDI) (100/6 μg, 2 puffs bid) for 2 weeks (visit 2); formoterol alone for 4 weeks (visit 3), and beclomethasone/formoterol combination for 4 weeks (visit 4). Pairwise PLS showed breath analysis with combined e-noses discriminates between visit 1 and 4 (accuracy: 90.3%; 34 discriminant sensors). EBC NMR-based metabolomics (accuracy: 72%) identified formate as the most discriminant metabolite. No between-visit difference in sputum inflammatory cells or FENO was observed. Pre-bronchodilator FEF25%-75% was lower (P=0.028) after formoterol monotherapy. Breathomics techniques, applied to gaseous/volatile phase vs aerosol particles, classify two types of ICS/LABA combination treatments. Further research is required to clarify possible therapeutic implications of these findings.