Lung cancer screening and the effects of competing causes of death in the ACRIN-NLST sub-study

Abstract

Background: CT screening for lung cancer is limited to older heavy smokers, in whom the risk of death from all the complications of smoking is high relative to the unscreened population. We have shown that airflow limitation is a marker of premature death from all causes and affects about 35% of National Lung Screening Trial (NLST) participants. The role of competing causes of death, relative to lung function, has not been examined in the NLST. Using 10,054 subjects from the ACRIN-NLST sub-study, this study aimed to compare disease-specific mortality according to the presence of airflow limitation to better understand its relationship with dying from lung cancer and dying from other smoking-related diseases. Methods: We stratified 10,054 NLST participants according to their FEV1%predicted values and compared rates of lung cancer, lung cancer death, and death from other causes, after a mean of 6.4 years of follow-up. Findings: There were 699 deaths; 189 (27%) attributed to lung cancer, 166 (24%) to cardiovascular disease, 61 (9%) to respiratory disease, and 150 (21%) to other cancers. After stratification according to baseline FEV1%predicted, we found a strong inverse relationship between FEV1%predicted and developing lung cancer as previously shown. We also found a strong inverse relationship between FEV1%predicted and dying of lung cancer and dying of disease other than lung cancer. As the FEV1%predicted decreased, the effect on increasing death from lung cancer and non-lung cancer causes diverged, favouring greater non-lung cancer deaths. Interpretation: In a CT screening study, where smokers were followed for a mean of 6.4 years, worsening lung function was associated with a greater tendency to die of non-lung cancer causes. We suggest smokers at the greatest risk of lung cancer (worse lung function) are at greatest risk of dying of diseases other than lung cancer, creating an important competing cause of death effect in screening. Funding: This study was supported by funding from Johnson and Johnson and grants from the National Cancer Institute, NIH, USA. Conflicts of interest: RPY is a shareholder of Synergenz Bioscience, which holds patents for gene-based risk prediction in COPD and lung cancer. All other authors declare that they have no conflicts of interest.