Longitudinal study to assess impact of smoking at diagnosis and quitting on 1-year survival for people with non-small cell lung cancer

Objectives: To update the prevalence of smoking in people as they were diagnosed with non-small cell lung cancer (NSCLC) and to see whether smoking status at baseline and quitting are independently associated with 1-year survival. Design: A real-world cohort study following patients from diagnosis for up to 1 year or until death. Setting: UK multi-centre study (28 sites) based in secondary and primary care. Participants: 1124 patients with newly diagnosed NSCLC between 2010-2016. Main outcome measures: Smoking status was validated at diagnosis and at every routine and emergency hospital visit. Cancer treatments were offered according to local multi-disciplinary team decisions following UK guidelines and smoking cessation treatments offered according to local practice /availability. Survival analysis and Cox Proportional Hazards Modelling examined the associations of a) smoking at baseline and b) quitting smoking, on survival at 1 year. Results: 77% of never smokers, 60% of ex-smokers and 57% of current smokers, were alive at 1 year (p = 0.01). After adjusting for age, stage, EGOG, surgery and gender, ex smokers (adjusted HR 1.96, 95% CI 1.16-2.31) and current smokers (aHR 2.04, 1.19-3.48) were both more likely to die within one year. 23% of smokers with NSCLC quit within 3 months of diagnosis. At 1 year, 69% of those who quit were alive versus 53% of those who continued to smoke (p < 0.01). After adjusting the risk of dying was lower (aHR 0.75), in those who quit smoking, although this was not statistically significant (p = 0.23). Conclusions: This is the largest prospective study that validates smoking in NSCLC; it shows a third of people are smoking at the time of diagnosis. Smokers have lower 12-month survival than never and ex-smokers. Quitting smoking was associated with 25% reduction in mortality which may be clinically important although not statistically significant, after adjusting for other factors.