Relationships between emphysema and airways metrics at High-Resolution Computed Tomography (HRCT) and ventilatory response to exercise in mild to moderate COPD patients

Background and purpose In Chronic Obstructive Pulmonary Disease (COPD) patients, the opportunity to carry out a thoracic high-resolution CT (HRCT) scan and to perform an incremental cardiopulmonary exercise test (CPET) increases the possibility to identify the different clinical features of disease. The aim of our study was to evaluate the relationships between HRCT metrics (on emphysema by low attenuation areas-LAA% and airways by wall area-WA%) and CPET variables related to the dynamic response to exercise in terms of elastic balance (Δ rest-to-peak IC/TLC) and ventilation capacity for carbon dioxide output (VE/VCO2 slope and VE/VCO2 intercept). Methods We prospectively enrolled COPD outpatients from the University Hospital of Parma. Data on anthropometrics characteristics, lung function, HRCT (LAA% and WA%) and CPET (Δ rest-to-peak IC/TLC, VE/VCO2 slope and VE/VCO2 intercept) were recorded. Results Fifty-one mild to moderate COPD patients (66% males; median age 70 y; mean FEV1 56% of pred.) were enrolled in the study. LAA% demonstrated a significant correlation with Δ rest-to-peak IC/TLC and VE/VCO2 slope (r = 0.405, p = 0.005 and r = 0.453, p = 0.001, respectively), while WA% with VE/VCO2 slope (r = -0.333, p = 0.020). In multivariate regression models, after adjustment for oxygen uptake (peak VO2) and physical capacity (peak workload), LAA was the only independent predictor of Δ rest-to-peak IC/TLC (β 0.774, SE 0.334, p = 0.025) and VE/VCO2 slope (β 0.155, SE 0.053, p = 0.005 and β 0.305, SE 0.123, p = 0.018, respectively). VE/VCO2 intercept was instead predicted from FEV1 only (β -0.097, SE 0.042, p = 0.027). Conclusion In mild to moderate COPD patients, emphysema (LAA) and airways metrics (WA) have close relationships with the different characteristics of ventilatory response to exercise. In particular, we were able to show that LAA is an independent predictor of exercise-induced Δ rest-to-peak IC/TLC and VE/VCO2 slope