Tobacco smoking and death from prostate cancer in US veterans

Abstract

Background Cigarette smoking is a risk factor for mortality in several genitourinary cancers, likely due to accumulation of carcinogens in urine. However, in prostate cancer (PC) the link has been less studied. We evaluated differences in prostate cancer-specific mortality (PCSM) between current smokers, past smokers, and never smokers diagnosed with PC. Methods This was a retrospective cohort study of PCSM in men diagnosed with PC between 2000 and 2015 treated in the US Veterans Affairs health care system, using competing risk regression analyses. Results The cohort included 73,668 men (current smokers: 22,608 (30.7%), past smokers: 23,695 (32.1%), and never smokers: 27,365 (37.1%)). Median follow-up was 5.9 years. Current smoker patients were younger at presentation (median age current: 63, never: 66; p \u2009<\u20090.001), and had more advanced disease stage (stage IV disease current: 5.3%, never: 4.3%; p \u2009<\u20090.04). The 10-year incidence of PCSM was 5.2%, 4.8%, and 4.5% for current, past, and never smokers, respectively. On competing risk regression, current smoking was associated with increased PCSM (subdistribution hazard ratio: 1.14, 95% confidence interval: (1.05–1.24), p \u2009=\u20090.002), whereas past smoking was not. Hierarchical regression suggests that this increased risk was partially attributable to tumor characteristics. Conclusions Smoking at the time of diagnosis is associated with a higher risk of dying from PC as well as other causes of death. In contrast, past smoking was not associated with PCSM suggesting that smoking may be a modifiable risk factor. PC diagnosis may be an important opportunity to discuss smoking cessation.