The effect of smoking on prostate cancer survival: a cohort analysis in Barcelona.

Abstract

The aim of this study was to analyse the effect of smoking on prostate cancer-specific mortality and all-cause mortality. A retrospective cohort study was conducted with 1109 patients with prostate cancer diagnosed from 1992 to 2008, identified through the Hospital del Mar Cancer Registry (Barcelona, Spain). Information on smoking habits was retrieved from clinical records and patients were classified into three categories: never smoker, exsmoker and current smoker. Patients were followed up until December 2011. Survival curves were plotted using Kaplan-Meier methods. Cox models were used to estimate hazard ratios and 95% confidence intervals. Median age at diagnosis was 70.6 years and 16.7% of patients had stage IV tumours. During the follow-up period, 466 deaths occurred, 36.1% of them being specifically due to prostate cancer. The median follow-up time of the censored patients was 5.8 years. There was a significant difference in disease-specific survival between never smokers, exsmokers and current smokers (P=0.0001). Current smokers presented a worse 5-year survival rate (82.9%) compared with exsmokers (88.9%) and never smokers (89.6%). In the multivariate analysis, after adjusting for age, disease stage, Gleason score and prostate-specific antigen, the hazard ratio for smokers was 1.80 (95% confidence interval: 1.04-3.13) compared with never smokers. In the exsmokers group the risk for prostate cancer-specific mortality was very similar to that of never smokers. However, the statistical difference disappeared when we stratified by stage (I-III and IV). In conclusion, smoking was identified as an independent and negative prognostic factor for prostate cancer-specific and all-cause mortality. These findings suggest that smoking-cessation programmes could be beneficial for prostate cancer patients.