Yang S, et al. Mol Oncol 2020.
Few studies have examined the association between smoking status (including former smokers) at diagnosis and overall survival among cancer patients. We aimed to assess the benefits of quitting smoking on cancer prognosis in cohorts of cancer patient smokers obtained from the Cancer Genome Atlas (TCGA) database. Hazard ratios (HR) were calculated to evaluate smoking behavior at cancer diagnosis (reformed smokers vs. current smokers) in association with overall survival using multivariate-adjusted
Cox regressions analysis. According to our analyses, quitting smoking was the independent protective factor for overall survival in lung squamous cell carcinoma (LUSC) (HR=0.67, 95%CI=0.48-0.94). Comprehensive analysis of multi-component data across reformed and current smokers identified a total of 85 differential expressed genes (DEGs) affected by different modes of genetic and epigenetic regulation, potentially representing cancer drivers in smokers. Moreover, we provided a smoking-associated gene expression signature which could evaluate the true effect on prognosis with high power (HR=1.70, 95%CI=1.19-2.43, AUC=0.65, 0.67 and 0.70 for 2, 3 and 5-year survival, respectively). This signature was also applicable in other smoking-related cancers, including bladder urothelial carcinoma (HR=1.70, 95%CI=1.01-2.88), cervical carcinoma (HR=5.69, 95%CI=1.37-23.69), head and neck squamous cell carcinoma (HR=1.97, 95%CI=1.41-2.76), lung adenocarcinoma (HR=1.73, 95%CI=1.16-2.57), and pancreatic adenocarcinoma (HR=4.28, 95%CI=1.47-12.47). In conclusion, this study demonstrates that quitting smoking at diagnosis decreases risk of death in cancer patients. We also provide a smoking-associated gene expression signature to evaluate the effect of smoking on survival. Lastly, we suggest that smoking cessation could comprise a part of cancer treatment to improve survival rates of cancer patients.