Does smoking status influence the prognosis of bladder cancer? A systematic review.

OBJECTIVE: To summarize, in a systematic review, the evidence for the effect of stopping smoking on recurrence, cancer-specific and all cause-mortality among smokers with newly diagnosed bladder cancer.
MATERIALS AND METHODS: Two electronic databases and the reference lists of identified primary studies and reviews were searched. Studies were included if a hazard ratio and its confidence intervals could be extracted. A predefined set of study characteristics was extracted which defined whether studies were giving valid prognostic data on the effects of smoking in reasonably homogenous cohorts. The results of studies were synthesized qualitatively.
RESULTS: Fifteen relevant studies were identified; former and current smokers were combined in many studies. Many studies produced information on prognosis that was confounded by the mixing of incident and prevalent cases. Only three studies examined the influence of smoking on prognosis in only incident cases, most of whom had superficial disease. Of these, only one was of high quality. These three studies and the other 12 showed suggestive evidence that continued smoking or a lifetime of smoking constitutes a moderate risk factor for recurrence and death, and that stopping smoking could favourably change this. However, the evidence base for this is weak because of the methodological shortcomings and because most studies' results were not statistically significant. A life-table model showed that if stopping smoking altered the prognosis, the size of the benefit would be clinically worthwhile.
CONCLUSION: There is suggestive evidence that stopping smoking might favourably alter the course of bladder cancer, but this is insufficient for clinicians to inform patients that doing so will improve their prognosis, and for providing specialized services to assist in stopping smoking to patients with bladder cancer.