Lifetime Smoking and its Impact on Risk for Lung Cancer and Cardiovascular Disease: Results from the Framingham Heart Study

Abstract

Among Framingham Heart Study (FHS) participants who smoked at least 20 pack-years, former heavy smokers’ risk of ASCVD and lung cancer may persist for up to 16 and beyond 25 years since quitting, respectively, compared to those who never smoke. Currently, American Heart Association/American College of Cardiology (AHA/ACC) ASCVD risk calculators assume former smokers quit >5 years have the same ASCVD risk as never smokers. Thus, former smokers’ ASCVD risk may be underestimated by such tools. Furthermore, despite their elevated risk, most heavy ever smokers never develop lung cancer while some never smokers do develop lung cancer; this may suggest a gene-by-smoking interaction. To fill these gaps in knowledge, we used longitudinal data from FHS to: 1 – assess the added value of incorporating pack-years smoked and years since quitting into ASCVD risk prediction models; and 2 – determine whether there is a gene-by-smoking interaction on lung cancer risk and explore whether a polygenic risk score (PRS) could be used to distinguish which ever smokers develop lung cancer outside of the current screening guidelines. Adding pack-years smoked and years since quitting to the current AHA/ACC ASCVD risk calculator had significant but modest continuous net reclassification improvement (NRI(>0)) and relative integrated discrimination improvement (rIDI) values in both sexes (Men: NRI(>0)=0.23, rIDI=0.19; Women: NRI(>0)=0.34, rIDI=0.11). In lung cancer analyses, we observed significant effect modification between the PRS and pack-years smoked on lung cancer risk such that the relative effect of pack-years smoked decreased with increasing PRS. We observed no significant association between the PRS and meeting lung cancer screening eligibility at time of diagnosis. Our results have important implications for clinical practice and suggest that pack-years smoked and years since quitting smoking should be incorporated into 10-year ASCVD risk estimation and that future research should focus on the impact of genetic burden and its interaction with pack-years smoked when assessing lung cancer risk.