A31-14 COPD and Pre-COPD: Where an Increase in Lung Cancer Risk Co-exists With an Increased Risk of Cardiovascular Death?

Abstract Rationale Lung Cancer (LC) screening can be optimised by screening high risk individuals based on age and smoking history. Those at highest risk can be better targeted through risk models based on other clinical risk variables including gender, BMI, education and comorbidity (e.g., PLCOM2012 model). However, there is growing concern that comorbidity may compromise LC screening outcomes due to competing mortality, where high risk screening participants die prematurely of non-LC causes. Aim This study aimed to better understand the relationship between LC risk and death from non-LC causes, specifically how cardiopulmonary comorbidity contributes to cause-specific mortality in a group of high risk ever smokers undergoing LC screening. Methods In a secondary analysis from the National Lung Screening Trial (N = 18,463), we used pre-bronchodilator spirometry and self-reported comorbidity to assign subjects into chronic obstructive pulmonary disease (COPD)-related subgroups (Healthy, PRISm, GOLD 0, GOLD 1, GOLD 2 and GOLD 3-4). In adjusted models, we compared the contribution these COPD-related phenotypes made to developing LC (using logistic regression) and risk of dying of LC or a competing death (using Competing Risk Cox-proportional models). We also compared the predictive utility of the PLCOM2012 in predicting mortality from lung cancer, cardiovascular and respiratory causes. Results Relative to the “healthy” group, the LC deaths increased 2-fold in those with pre-COPD and 2-4 fold in those with increasingly worse airflow limitation (GOLD 1,2 and 3-4)(Figure 1). Increases in LC mortality were associated with increases in cardiovascular mortality, that collectively accounted for over 50% of all deaths. In adjusted logistic regression, although pre-COPD (PRISm and GOLD 0) and COPD (GOLD 1, 2, 3-4) independently conferred an increased risk of developing and dying of LC, PRISm and GOLD 2-4 also independently increased the risk for dying of cardiovascular and respiratory deaths, GOLD 0 increased the risk for dying of respiratory disease along with PRISm and GOLD 2-4. Diabetes was associated with lung cancer and cardiovascular mortality. The predictive utility of the PLCOM2012 model for lung cancer deaths was comparable to that for respiratory deaths and slightly lower for CVS deaths. Conclusion After extensive adjustment for shared risk factors (age, smoking, gender, BMI and cardiovascular history), respiratory comorbidity (Pre-COPD and COPD), that conferred a greater risk of dying of LC also confers a greater risk of cardiovascular deaths. We conclude that enriching for lung cancer risk (or death) enriches for cardiovascular-related death that may undermine LC screening benefits. This abstract is funded by: None