BackgroundComputed tomography (CT) phenotypic characterization helps in understanding the clinical diversity of chronic obstructive pulmonary disease (COPD) patients, but its clinical relevance and its relationship with functional features are not clarified. Volumetric capnography (VC) uses the principle of gas washout and analyzes the pattern of CO2 elimination as a function of expired volume. The main variables analyzed were end-tidal concentration of carbon dioxide (ETCO2), Slope of phase 2 (Slp2), and Slope of phase 3 (Slp3) of capnogram, the curve which represents the total amount of CO2 eliminated by the lungs during each breath.ObjectiveTo investigate, in a group of patients with severe COPD, if the phenotypic analysis by CT could identify different subsets of patients, and if there was an association of CT findings and functional variables.Subjects and methodsSixty-five patients with COPD Gold III–IV were admitted for clinical evaluation, high-resolution CT, and functional evaluation (spirometry, 6-minute walk test [6MWT], and VC). The presence and profusion of tomography findings were evaluated, and later, the patients were identified as having emphysema (EMP) or airway disease (AWD) phenotype. EMP and AWD groups were compared; tomography findings scores were evaluated versus spirometric, 6MWT, and VC variables.ResultsBronchiectasis was found in 33.8% and peribronchial thickening in 69.2% of the 65 patients. Structural findings of airways had no significant correlation with spirometric variables. Air trapping and EMP were strongly correlated with VC variables, but in opposite directions. There was some overlap between the EMP and AWD groups, but EMP patients had signicantly lower body mass index, worse obstruction, and shorter walked distance on 6MWT. Concerning VC, EMP patients had signicantly lower ETCO2, Slp2 and Slp3. Increases in Slp3 characterize heterogeneous involvement of the distal air spaces, as in AWD.ConclusionVisual assessment and phenotyping of CT in COPD patients is feasible and may help identify functional and clinically different subsets of patients. VC may provide useful information about the heterogeneous involvement of lung structures in COPD.