Abstract
Purpose Prolonged warfarin use may decrease the risk of prostate cancer. We aimed to assess the effect of warfarin on histological grade and clinical stage of prostate cancer at diagnosis.
Methods We carried out a retrospective population-based cohort study of men older than 50 years of age diagnosed with prostate cancer between 1985 and 2002 and registered with the Saskatchewan Cancer Registry. We compared a composite score of histological grade and clinical stage of prostate cancer at diagnosis according to warfarin use in the 5 years preceding the diagnosis of prostate cancer.
Results Compared with non-users, men with at least 2 years of cumulative warfarin use in the 5 year period preceding the diagnosis of prostate cancer were at a lower risk of a poor prognosis composite score at the time of their prostate cancer diagnosis (OR 0.40, 95%CI (0.19-0.83)), and when intermediate and poor prognosis scores were combined, a similar estimate of association was observed (OR 0.55, 95%CI (0.33-0.91)), adjusted for age at diagnosis and year of diagnosis. However, an increased risk of poor prognosis disease was observed with 4 years of cumulative warfarin use compared to never use (OR 2.2, 95%CI (1.03-4.81)).
Conclusions There is a suggestion that at least 2 years of warfarin use is associated with a more favourable prognosis but that extended duration of use beyond 2 years may be associated with poor prognosis disease. Further investigation with a more complete assessment of confounders and that addresses potential detection biases is warranted. Copyright (C) 2010 John Wiley & Sons, Ltd.