Abstract
OBJECTIVE: Patients eligible for lung cancer screening are at risk for structural lung abnormalities, especially emphysema. Our aim was to assess the association between emphysema as determined by radiologist and quantitative CT (QCT) analysis, and the presence of spirometric abnormalities in an urban largely minority population undergoing lung cancer screening. DESIGN: Retrospective cohort study. SETTING: Centralised lung cancer screening programme in an urban academic medical centre in Philadelphia, Pennsylvania, USA. PARTICIPANTS: 2111 individuals who underwent low-dose CT (LDCT) from October 2021 to October 2022 had CT scans interpreted by radiologists for the presence of emphysema, 629 of whom underwent spirometry. 181 patients underwent AI-assisted QCT analysis for emphysema at -950 and -910 HU, all of whom underwent spirometry. PRIMARY AND SECONDARY OUTCOME MEASURES: Diagnostic tests for presence of emphysema as defined by radiologist and by AI-assisted QCT analysis for emphysema and the association with spirometric obstruction. RESULTS: A total of 2111 individuals underwent LDCT from October 2021 to October 2022. 323 (51.4%) individuals with radiology-determined emphysema had airflow obstruction. 108 (17.2%) individuals without emphysema had no airflow obstruction. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of radiology detected emphysema for airflow obstruction are 81.3%, 53.4%, 72.2%, 59.34%, respectively. In the Temple Healthy Chest Initiative (THCI) QCT subcohort of 181 individuals who underwent spirometry and low-attenuation area (%LAA)-910 HU, 91 (50%) individuals had %LAA-910 HU≥6% and airflow obstruction. 33 (18%) individuals with %LAA-910<6% had no airflow obstruction. Sensitivity, specificity, PPV and NPV for the presence of airflow obstruction based on %LAA-910 HU≥6% were 63%, 49%, 73% and 59%, respectively. CONCLUSIONS: Data from the THCI demonstrates high rates of emphysema as detected by radiologic qualitative interpretation and quantitative analysis. Using a cut-off of ≥6% for LAA at -910 HU, there is a high sensitivity and PPV for associated spirometric obstruction and thus should drive providers to obtain prebronchodilator spirometry to detect chronic obstructive pulmonary disease at an earlier stage.