Background: Lung cancer risk attributable to smoking is dose dependent, yet few studies examining a polygenic risk score (PRS) by smoking interaction have included comprehensive lifetime pack-years smoked.
Methods: We analyzed data from participants of European ancestry in the Framingham Heart Study Original (n = 454) and Offspring (n = 2,470) cohorts enrolled in 1954 and 1971, respectively, and followed through 2018. We built a PRS for lung cancer using participant genotyping data and genome-wide association study summary statistics from a recent study in the OncoArray Consortium. We used Cox proportional hazards regression models to assess risk and the interaction between pack-years smoked and genetic risk for lung cancer adjusting for European ancestry, age, sex, and education.
Results: We observed a significant submultiplicative interaction between pack-years and PRS on lung cancer risk (P = 0.09). Thus, the relative risk associated with each additional 10 pack-years smoked decreased with increasing genetic risk (HR = 1.56 at one SD below mean PRS, HR = 1.48 at mean PRS, and HR = 1.40 at one SD above mean PRS). Similarly, lung cancer risk per SD increase in the PRS was highest among those who had never smoked (HR = 1.55) and decreased with heavier smoking (HR = 1.32 at 30 pack-years).
Conclusions: These results suggest the presence of a submultiplicative interaction between pack-years and genetics on lung cancer risk, consistent with recent findings. Both smoking and genetics were significantly associated with lung cancer risk.
Impact: These results underscore the contributions of genetics and smoking on lung cancer risk and highlight the negative impact of continued smoking regardless of genetic risk.
©2024 The Authors; Published by the American Association for Cancer Research.