Smoking history, eligibility for lung cancer screening and risk of death by lung cancer or other causes -- a longitudinal, time-variable analysis of the EPIC-Heidelberg cohort

Abstract Background Eligibility criteria for lung cancer (LC) screening aim to identify long-term smokers who have exceeded a minimum-risk threshold of having LC, while still having sufficiently long remaining life expectancy. For individuals who have once met eligibility criteria (i.e., reached sufficiently high LC risk), a question is whether longer-term smoking cessation and corresponding improvement in other-cause mortality risk could justify a higher maximum screening age for former smokers than for continuing smokers. Methods We performed time-varying Cox models in the EPIC-Heidelberg cohort (N = 24,715), using 3-yearly questionnaire data collected between 1994 and 2014 to estimate hazard ratios (HRs) and incidence rates for death by LC or other causes, in relation to age-specific smoking status and eligibility by German LC screening criteria (LCSC). Results Depending on age, models showed up to 3-fold higher risk of other-cause mortality for recent and LCSC-eligible smokers, relative to never smokers. Among former smokers, those who quit before age 30 or 40 showed no significant difference in other-cause mortality compared with never-smokers, in men and women, respectively. However, at higher quitting ages HRs increased up to about 1.9 for men who stopped at age ≥60, or 1.5 in women who stopped at age 50-<60. Also, depending on age, former smokers who had once met the LCSC, but then quit for >10 years (thereby losing formal screening eligibility), showed HRs of about 1.5–3.0 for other-cause mortality in men and 1.3–1.9 in women. In both sexes, the absolute incidence rate for other-cause mortality amongst past-eligible smokers age 75-<80 was similar to that for current-eligible smokers of the younger, 70-<75-year age group. For LC, both current and former smoking were associated with persistently increased HRs, even after long-term cessation. Conclusions For smokers who once met LCSC, but then quit for >10 years the risk of other-causes mortality remains elevated, but less so than for continuing smokers, which may argue for a moderate extension of the maximum age limit for LC screening. Larger studies will be needed to obtain more precise risk estimates.