The prevalence and physiological impacts of centrilobular and paraseptal emphysema on computed tomography in smokers with preserved ratio impaired spirometry.

Centrilobular emphysema (CLE) and paraseptal emphysema (PSE) are observed in smokers with preserved ratio impaired spirometry (PRISm, defined as the ratio of forced expiratory volume in 1 s (FEV) to forced vital capacity (FVC) ≥0.7 and FEV <80%), but their prevalence and physiological impacts remain unestablished. This multicentre study aimed to investigate its prevalence and to test whether emphysema subtypes are differently associated with physiological impairments in smokers with PRISm. Both never- and ever-smokers aged ≥40 years who underwent computed tomography (CT) for lung cancer screening and spirometry were retrospectively and consecutively enrolled at three hospitals and a clinic. Emphysema subtypes were visually classified according to the Fleischner system. Air-trapping was assessed as the ratio of FVC to total lung capacity on CT (TLC). In 1046 never-smokers and 772 smokers with ≥10 pack-years, the prevalence of PRISm was 8.2% and 11.3%, respectively. The prevalence of PSE and CLE in smokers with PRISm was comparable to that in smokers with normal spirometry (PSE 43.7% 36.2%, p=1.00; CLE 46.0% 31.8%, p=0.21), but higher than that in never-smokers with PRISm (PSE 43.7% 1.2%, p<0.01; CLE 46% 4.7%, p<0.01) and lower than that in smokers with airflow limitation (PSE 43.7% 71.0%, p<0.01; CLE 46% 79.3%, p<0.01). The presence of CLE, but not PSE, was independently associated with reduced FVC/TLC in smokers with PRISm. Both PSE and CLE were common, but only CLE was associated with air-trapping in smokers with PRISm, suggesting different physiological roles of these emphysema subtypes.