EUF Vol. 1 No. 1

34 96

with PCa risk (p = 0.09). No significant, clear pattern of
association was observed between smoking duration and
PCa risk. Previous smoking showed no association (RR: 1.00;
95% CI, 0.95–1.06), but ever smoking showed an inverse
association (RR: 0.94; 95% CI, 0.90–0.98) with incident PCa;
however, heterogeneity in results for both groups was high
(previous smoking: I2 = 61%; p < 0.001; ever smoking:
I2 = 68%; p < 0.001). It is unlikely that the difference in
pattern of association over time is related solely to
differences in the quality of studies. Some earlier studies
were large, well-conducted studies. One possible explana-
tion, suggested by the authors, is that smoking may reduce
the risk of indolent nonaggressive cancers, which predomi-
nate among cancers detected in more recent years, while
promoting more aggressive cancers Alternatively,
smoking has been linked with lower risk of screening and
poor compliance with prostate biopsy . A recent
analysis of the Reduction by Dutasteride of Prostate Cancer
Events (REDUCE) study, in which men with a negative
baseline biopsy and elevated prostate-specific antigen
(PSA) were randomized to dutasteride or placebo and were
required to under biopsy at 2 and 4 yr, found that smokers
were less compliant . On the 2-yr biopsy, smokers had
more high-grade disease; when considering the whole
REDUCE study and the fewer biopsies performed, this effect
was lost. On a population level, perhaps smokers are less
likely to be screened, resulting in the detection of fewer
nonaggressive PCa screen-detected cancers and making
smoking appear ‘‘protective’’ in more recent years.
Regarding PCa mortality (as opposed to PCa incidence),
the data were more clear. A significant 14% increased risk of
PCa death associated with current smoking was reported in
that meta-analysis . The highest categories of smoking
were associated with 24–30% increased risk . Results
of several prospective studies were published recently and
included in a meta-analysis by Islami et al The recent
meta-analysis observed a robust association between ciga-
rette smoking and PCa death: It was observed in analyses of
current, former, and ever use and inmeta–regressionmodels,
suggesting a dose-response association, and persisting in
subgroup analyses includingwhen stratified by geography or
study completion time . Current cigarette smoking at
baseline was associated with an increased risk of death from
PCa (RR: 1.24; 95% CI, 1.18–1.31) with little heterogeneity in
results (I2 = 1%; p = 0.45). In meta–regression models, the
amount of cigarette smoking at baseline (cigarettes per day)
showed a dose-response association with PCa death
(p = 0.02; 20 cigarettes per day, RR: 1.20). The RR for the
association between previous cigarette smoking at baseline
and PCa mortality was 1.06 (95% CI, 1.00-1.13) with little
heterogeneity (I2 = 0%; p = 0.62). The RR for the association
between ever cigarette smoking and PCa mortality was 1.18
(95% CI, 1.11–1.24) but with moderate, statistically signifi-
cant heterogeneity (I2 = 36%; p = 0.04).
Limitations of the available evidence should be consid-
ered. Only a few of the studies included in the two
meta-analyses provided information about PCa screening in
their study populations, probably because this information
was not available from most cancer registers, which were
the main sources of outcome measures . The few
papers that did provide this information suggest that the
association between smoking and PCa death may be slightly
stronger in those with no screening compared with those
with PCa screening . It is also important to
Smoking
Inflammation Carcinogenic substance Hormone changes
Increased total and free
testosterone,
increased estradiol
Gene polymorphism
mutation
p53
Cytochrome P450
Glutathione-S-transferases
Hemeoxygenase 1
Aromatic hydrocarbons
cadmium
Cytokine release
(IL-6, IL-8, IL-18, TGF-α)
Prostate cancer
Fig. 2 – Biological hypothesis for prostate cancer development and smoking.
E U R O P E A N U R O L O G Y F O C U S 1 ( 2 0 1 5) 2 8 – 3 8 34