Prostate Cancer Progression: The Smoking Gun?

Abstract

In this issue of European Urology, Rieken et al [1] provide further evidence supporting an association of smoking with risk of clinical outcomes in men with prostate cancer. In prior studies, smoking has been associated with biochemical recurrence (BCR), metastasis, hormone refractory prostate cancer, and prostate cancer mortality in patients receiving one type of treatment [2–6] and in studies adjusted for the type of treatment [7]. With 22 yr of followup among 5366 men with prostate cancer in the Health Professionals Follow-Up Study, we showed that current smoking was associated with a 61% increased risk of BCR as well as a 61% increased risk of prostate cancer mortality [7]. As with the current study, we did not observe an association between cumulative dose measured by packyears for BCR, and also like Rieken et al [1], we found that smokers who had quit for 10 yr or more had similar risks as people who had never smoked, regardless of pack-years. The Rieken et al [1] study combined data from6538 patients treatedwith radical prostatectomy fromsixUSandEuropean sites, with a median follow-up of 28 months in those without BCR. Their findings support a positive association between current smokers and recent quitters with BCR. Smoking is one of a growing number of risk factors that have been linked to a higher risk of prostate cancer progression and lethality [8], despite a lack of association with the overall incidence of prostate cancer. Rieken et al [1] note that smoking has been associated with an overall decreased riskofprostate cancer incidence [9], particularly in more recent studies conducted after prostate-specific antigen (PSA) screening became common practice. They speculate that ‘‘smoking may reduce the risk of indolent cancers,’’ but a far more plausible explanation for this association is that it represents an artifact of differential screening practices. Smokers are substantially less likely to have intensive PSA screening and biopsies and are therefore less likely to be diagnosed with indolent disease. It is important to be aware of the enormous impact of PSA screening in analyses of total prostate cancer. Due to the preponderance of indolent disease, research should focus mainly on the clinically important endpoint of lethal prostate cancer. In contrast to the lack of association between smoking and the risk of incident[1_TD$DIFF] prostate cancer, the link between smoking and poor prostate-specific outcomes is increasingly well-established. Studies from before the PSA era, as well as more recent studies that have adjusted for the intensity of PSA screening, suggest that the link with lethal prostate cancer is not due simply to less PSA screening among smokers. Instead, a true biological cause and effect relationship seems likely. Several potential mechanisms have been proposed, including that the induction of heme oxygenase-1 expression leads to an increased expression of vascular endothelial growth factor, which is discussed by Rieken et al [1]. The timing for the effect of smoking—acting in the several years prior to diagnosis—is consistent with such a mechanism. The magnitude of the increase in risk for BCR in this study—an approximate doubling in risk—is quite similar to what we previously reported for fatal prostate cancer. This similarity of relative risk estimates is of interest, because only a minority of men who experience BCR go on to develop metastatic disease or die from prostate cancer. If the association of smoking with BCRwas simply a reflection of its link to potentially fatal disease, one would expect the relative risk estimates to be substantially smaller for BCR as EU RO P E AN URO LOGY 6 8 ( 2 0 1 5 ) 9 5 7 – 9 5 8