You are here
The Role of Tobacco Smoke in Bladder and Kidney Carcinogenesis: A Comparison of Exposures and Meta-analysis of Incidence and Mortality Risks
European Urology, In Press, Corrected Proof, Available online 3 July 2015, January 1970
Tobacco smoke includes a mix of carcinogens implicated in the etiology of bladder cancer (BC) and renal cell cancer (RCC).
We reviewed the impact of tobacco exposure on BCC and RCC incidence and mortality, and whether smoking cessation decreases the risk.
A systematic review of original articles in English was performed in August 2013. Meta-analysis of risks was performed using adjusted risk ratios where available. Publication bias was assessed using Begg and Egger tests.
We identified 2683 papers, of which 112 fulfilled our inclusion criteria, of which 88 studies investigated BC and 24 investigated RCC. The pooled relative risk (RR) of BC incidence was 2.62 (95% confidence interval [CI] 2.43–2.83) for all smokers, 3.49 (3.13–3.88) for current smokers, and 2.07 (1.84–2.33) for former smokers. The corresponding pooled RR of BC disease-specific mortality (DSM) was 1.48 (1.20–1.84), 1.43 (1.20–1.71) and 1.24 (0.99–1.55). The pooled RR of RCC incidence was 1.30 (1.21–1.39) for all smokers, 1.33 (1.17–1.52) for current smokers, and 1.17 (1.09–1.24) for former smokers. The corresponding RCC DSM risk was 1.20 (1.02–1.41), 1.32 (1.08–1.62), and 1.01 (0.85–1.20).
We present an up-to-date review of tobacco smoking and BC and RCC incidence and mortality. Tobacco smoking significantly increases the risk of BC and RCC incidence. BC incidence and DSM risk are greatest in current smokers and lowest in former smokers, indicating that smoking cessation confers benefit. We found that secondhand smoke exposure is associated with a significant increase in BC risk.
Tobacco smoking affects the development and progression of bladder cancer and renal cell cancer. Smoking cessation reduces the risks of developing and dying from these common cancers. We quantify these risks using the most up-to-date results published in the literature.
Keywords: Bladder cancer, Kidney cancer, Tobacco smoking.
Tobacco smoke is the commonest human carcinogen. The World Health Organization estimates that in 2013 there were more than one billion smokers worldwide  and approximately six million people die each year from tobacco-related illnesses. These deaths include an estimated one million nonsmokers who obtained exposure indirectly from environmental tobacco smoke or secondhand smoking (SHS)  . The majority of smoking-related deaths occur because of cardiovascular and pulmonary diseases or malignancies. The risk of tobacco-related illnesses varies with the duration and intensity of smoking  , the type of tobacco and mode of administration, and an individual's ability to detoxify carcinogens. Tobacco can be consumed in a variety of forms such as smoking cigarettes, cigars, pipes, and shisha (a molasses-tobacco hybrid compound), chewing, and inhalation as snuff, and can be used in isolation or in combination with illicit drugs such as opium and marijuana  . Tobacco can be prepared via flue (blonde) or air curing (black). The latter is considered to be more carcinogenic to the urinary tract owing to its higher concentration of nitrosamines, biphenyls, and arylamines , , and . With regard to carcinogen detoxification, variations in the activity of N-acetyl-transferase 2 (NAT2) and glutathione S-transferase mu μ1 (GSTM1) because of polymorphisms appear to affect cancer risk from smoking  . It is also evident that tobacco smoke can induce changes in the DNA damage response machinery, which can additively or synergistically impair the host response to carcinogens  and .
Bladder cancer (BC) and renal cell cancer (RCC) are among the commonest smoking-related human malignancies. In 2013 there were an estimated 382 700 new cases of BC and 338 000 of RCC worldwide, with 143 000 and 150 300 resultant deaths, respectively  and . Both tumors are more common in males than females, reflecting the role of tobacco smoking, occupational carcinogen exposure, and lifestyle in their etiology. Tobacco smoke inhalation appears to be the commonest risk factor for BC, accounting for approximately 50% of BC cases  and 20–25% of RCC cases  . Further risk factors for RCC include obesity and hypertension. For both cancers, risk may be modified by genetic predisposition and interaction with further carcinogens  , and altering smoking exposure may change the natural history of the disease. For example, smoking cessation may reduce BC recurrence rates  , although conflicting data exist  and . Regardless of this contradiction, smoking-induced DNA damage (as detected in either blood or urine) reduces to normal levels after cessation  .
Here we present a systematic review of the literature and meta-analysis of the associations between smoking and both BC and RCC. We analyze both incidence and mortality, and specifically combine risks for SHS and non–smoking-related tobacco exposures. Owing to the causal relationship between active smoking and BC, there has been strong reason to suspect that SHS (also known as environmental tobacco smoke or passive smoking) has a role in carcinogenesis. The strength of this association has been emphasized by evidence that urinary levels of carcinogens are greater in subjects exposed to SHS than those not subjected to this exposure  .
2. Evidence acquisition
2.1. Systematic review
We searched PubMed in August 2013 for all original articles in English using the string terms “tobacco”, “smoking” AND “bladder cancer”, and “tobacco”, “smoking” AND “kidney cancer”. Articles were included in the meta-analysis if they met the following inclusion criteria: (i) case-control, cohort, or nested case-control studies published as original articles in English investigating the relationship between smoking and the risk of BC or RCC in humans; (ii) incidence or disease-specific mortality (DSM) as outcome; and (iii) odds ratio (OR), hazard ratio (HR), or relative risk (RR) estimates with 95% confidence intervals (CIs), or enough information to calculate them, reported. We excluded summary data (reviews) and reports not focusing on our research question or describing molecular effects in cell lines. In cases of multiple reports from the same series, we used the most recent one. Previous meta-analyses and systematic reviews were only included for discussion purposes when describing potential carcinogenic processes. We report our findings in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines  .
2.2. Data abstraction
From each study included in the meta-analysis, we extracted the first author's last name, publication year, country, study period, gender of study participants, cancer type (BC or RCC), number of cases and controls (for case-control or nested studies) or number of events and cohort size (for cohort studies), smoking status (all, former, or current), tobacco products (cigarettes, cigars, or pipes), SHS exposure, adjustment variables, and RRs or ORs with 95% CIs for each smoking status or tobacco product. If multiple RRs or ORs were presented in the original articles, we extracted the estimates from the maximally adjusted model to reduce the risk of possible unmeasured confounding  .
2.3. Statistical methods
Because cancer is a relatively rare outcome, we assumed that ORs, risk ratios, and rate ratios were all comparable estimates of the RR. To conduct the meta-analysis, measures of association and the corresponding CIs were translated into log(RR) values and their variances  .
BC and RCC incidence and DSM risks were computed separately. We used the maximum adjusted risk estimates when reported. We computed pooled RRs for BC and RCC incidence and DSM risks using a random effects model to take into account the heterogeneity between risk estimates  . We evaluated potential heterogeneity among studies using the Cochran Q statistic and I2, that is, the proportion of total variation contributed by between-study variance  .
To investigate potential sources of heterogeneity, we carried out stratified analyses according to study area (Europe, America, Asia, and Oceania), study design (case-control and cohort studies), and gender. We also tested whether the corresponding stratified pooled RR estimates differed significantly across the strata considered.
Stata statistical software (version 12.0, StataCorp LP, College Station, TX, USA) was used for statistical analysis.
3. Evidence synthesis
Our search identified 2683 reports (1237 BC, 225 RCC, 8 both, and 1213 unrelated cancers). All abstracts were read in full by one author (M.G.C.) before selection of 248 papers for extraction. From these full reports, we identified 112 articles (Supplementary Appendix 1) fulfilling our inclusion criteria for the meta-analysis (Supplementary Fig. 1). Outcomes for 51 230 BC cases and 64 602 controls, and for 15 265 RCC cases and 12 944 controls were included in the meta-analysis. Specifically, 107 papers included data on disease incidence or mortality in relation to cigarette smoking, eight papers concentrated on alternative means of tobacco exposure (eg, chewing), and five evaluated SHS (passive smoking). The majority of the reports focused on BC (79%).
3.1. BC incidence
We stratified BC risk according to current, former (no longer smoking at the time of interview), all (data for both current and former smokers, as well as data reported for ever smokers), and never smoker history ( Table 1 ). There were significant pooled RRs for BC incidence among all smokers of cigarettes (RR 2.62, 95% CI 2.43–2.83; I2 = 87.9%, p < 0.001), current smokers (RR 3.49, 95% CI 3.13–3.88; I2 = 80.7%, p < 0.001), and former smokers (RR 2.07, 95% CI 1.84–2.33; I2 = 83.4%, p < 0.001) when compared to never smokers. Current smokers had the greatest risk ( Fig. 1 ). When stratified by study design, a stronger association between smoking and BC risk was observed in case-control studies than in cohort studies ( Table 1 ). Publication bias for BC among all smokers was assessed using Begg (p = 0.04) and Egger (p = 0.2) tests. Visual inspection of a funnel plot could not rule out publication bias (Supplementary Fig. 2). We further stratified the data by gender and geographic region. Although males (RR 2.56, 95% CI 2.22–2.95; I2 = 92.2%, p < 0.001) had a slightly higher risk than females (RR 2.26, 95% CI 1.88–2.72; I2 = 85.1%, p < 0.001), pooled RR estimates did not differ across gender (p = 0.3). The majority of data came from studies based in North America (listed as Americas) and Europe. The highest pooled RR was observed in studies carried out in Europe (RR 2.98, 95% CI 2.65–3.36; I2 = 86.2%, p < 0.001; Table 3 ), although we did not see a difference across geographic region (p = 0.2). Among groups that used non-cigarette tobacco, cigar smoking (RR 1.62, 95% CI 1.18–2.22; I2 = 39.4%, p = 0.2) and pipe smoking (RR 1.49, 95% CI 1.18–1.88; I2 = 0.0%, p = 0.6) were both associated with significantly higher BC risk (Supplementary Table 2), although pooled RRs estimates were based on just a few studies. We did not observe a significant difference in pooled RR for smoking between non-cigarette tobacco products and cigarettes (p = 0.1).
|n a||PRR (95% CI)||I2, % (p value b )||n a||PRR (95% CI)||I2, % (p value b )|
|Case-control studies||117||2.78 (2.54–3.04)||88.9 (<0.001)||–||–||–|
|Cohort studies||36||2.02 (1.92–2.54)||83.7 (<0.001)||15||1.48 (1.20–1.84)||76.0 (<0.001)|
|Overall||153||2.62 (2.43–2.83)||87.9 (<0.001)||15||1.48 (1.20–1.84)||76.0 (<0.001)|
|Case-control studies||48||2.62 (2.28–3.03)||89.1 (<0.001)||0||–||–|
|Cohort studies||7||1.52 (1.25–1.83)||0.0 (0.7)||2||1.84 (0.74–4.59)||94.3 (<0.001)|
|Overall||55||2.46 (2.16–2.81)||87.9 (<0.001)||2||1.84 (0.74–4.59)||94.3 (<0.001)|
|Case-control studies||36||3.78 (3.33–4.28)||80.0 (<0.001)||0||–||–|
|Cohort studies||12||2.81 (2.30–3.43)||78.3 (<0.001)||7||1.43 (1.20–1.71)||11.1 (0.3)|
|Overall||48||3.49 (3.13–3.88)||80.7 (<0.001)||7||1.43 (1.20–1.71)||11.1 (0.3)|
|Case-control studies||33||2.06 (1.79–2.36)||78.8 (<0.001)||0||–||–|
|Cohort studies||17||2.08 (1.67–2.59)||87.3 (<0.001)||6||1.24 (0.99–1.55)||41.2 (0.1)|
|Overall||50||2.07 (1.84–2.33)||83.4 (<0.001)||6||1.24 (0.99–1.55)||41.2 (0.1)|
a Number of comparisons. Some studies include separate estimates for males and females and for smoking category.
b p value for heterogeneity.
3.2. BC mortality
BC mortality is less extensively reported in the literature. All smokers (RR 1.48, 95% CI 1.20–1.84; I2 = 76.0%, p < 0.001), current smokers (RR 1.43, 95% CI 1.20–1.71; I2 = 11.1%, p = 0.3) and former smokers (RR 1.24, 95% CI 0.99–1.55; I2 = 41.2%, p = 0.1) had a higher risk of BC mortality compared to never smokers ( Table 1 ). Cigar smoking had a nonsignificant higher mortality risk (data not shown). For current and former smokers, the Begg (p = 0.2 and 0.1) and Egger (p = 0.1 and 0.2) tests for publication bias confirmed that there was no significant publication bias. There were no significant differences by gender (p = 0.3), while a significant difference in pooled RRs emerged when we stratified by study geographic region (p = 0.019; Table 3 ).
3.3. RCC incidence
The risk of developing RCC was significantly higher for all smokers (RR 1.30, 95% CI 1.21–1.39; I2 = 55.6%, p < 0.001), current smokers (RR 1.33, 95% CI 1.17–1.52; I2 = 71.1%, p < 0.001), and former smokers (RR 1.17, 95% CI 1.09–1.24; I2 = 0.0%, p = 0.6; Table 2 ) compared to nonsmokers. Current smokers had the greatest risk ( Fig. 2 ). Begg and Egger tests for publication bias for all smokers (p = 0.9 and p = 0.5), current smokers (p = 0.3 and 0.9, Supplementary Fig. 3), and former smokers (p = 0.8 and 0.4) showed that there was no significant publication bias. A significant difference (p = 0.007) in pooled RRs emerged when we stratified by study geographic region; the greatest pooled RR for RCC was observed for Oceania (RR 1.74, 95% CI 1.14–2.66; I2 = 70.2%, p = 0.07) and the lowest for Europe (RR 1.02, 95% CI 0.91–1.12; I2 = 0.0%, p = 0.6). Stratification by gender revealed that males (RR 1.42, 95% CI 1.25–1.62; I2 = 55.0%, p = 0.001) had a slightly higher pooled RR for RCC than females (RR 1.32, 95% CI 1.16–1.51; I2 = 26.6%, p = 0.1), although the difference was not significant (p = 0.4). There were insufficient data on non-cigarette tobacco use and RCC risk.
|n a||PRR (95% CI)||I2, % (p value b )||n a||PRR (95% CI)||I2, % (p value b )|
|Case-control studies||46||1.29 (1.19–1.41)||62.7 (<0.001)||–||–||–|
|Cohort studies||18||1.31 (1.19–1.44)||18.1 (0.2)||8||1.20 (1.02–1.41)||51.6 (0.044)|
|Overall||64||1.30 (1.21–1.39)||55.6 (<0.001)||8||1.20 (1.02–1.41)||51.6 (0.044)|
|Case-control studies||14||1.45 (1.27–1.66)||45.4 (0.03)||0||–||–|
|Cohort studies||0||–||–||1||1.30 (0.92–1.84)||–|
|Overall||14||1.45 (1.27–1.66)||45.4 (0.03)||1||1.30 (0.92–1.84)||–|
|Case-control studies||17||1.33 (1.12–1.59)||76.0 (<0.001)||0||–||–|
|Cohort studies||8||1.33 (1.09–1.63)||57.2 (0.022)||4||1.32 (1.08–1.62)||25.8 (0.3)|
|Overall||25||1.33 (1.17–1.52)||71.1 (<0.001)||4||1.32 (1.08–1.62)||25.8 (0.3)|
|Case-control studies||15||1.13 (1.04–1.22)||5.1 (0.4)||0||–||–|
|Cohort studies||10||1.26 (1.12–1.43)||0.0 (0.9)||3||1.01 (0.85–1.20)||14.5 (0.3)|
|Overall||25||1.17 (1.09–1.24)||0.0 (0.6)||3||1.01 (0.85–1.20)||14.5 (0.3)|
a Number of comparisons. Some studies include separate estimates for males and females and for smoking category.
b p value for heterogeneity.
3.4. RCC mortality
The risk of death from RCC among tobacco users was elevated for all smokers (RR 1.20, 95% CI 1.02–1.41; I2 = 51.6%, p = 0.044), current smokers (RR 1.32, 95% CI 1.08–1.62; I2 = 25.8%, p = 0.3), and former smokers (RR 1.01, 95% CI 0.85–1.20; I2 = 14.5%, p = 0.3; Table 2 ). Stratification by geographic region revealed that the greatest RR for RCC was in the Americas, but the pooled RR did not differ (p = 0.8), although the numbers are small ( Table 3 ).
|n a||PRR (95% CI)||I2, % (p value b )||n a||PRR (95% CI)||I2, % (p value b )|
|Male||45||2.56 (2.22–2.95)||92.2 (<0.001)||5||1.92 (1.49–2.47)||0.0 (0.9)|
|Female||35||2.26 (1.88–2.72)||85.1 (<0.001)||2||1.30 (0.68–2.51)||0.0 (0.7)|
|Mixed||73||2.83 (2.56–3.14)||84.4 (<0.001)||8||1.35 (1.00–1.80)||86.4 (<0.001)|
|Europe||66||2.98 (2.65–3.36)||86.2 (<0.001)||5||2.26 (1.69–3.04)||32.4 (0.2)|
|Asia||16||2.26 (1.79–2.86)||83.3 (<0.001)||5||1.56 (1.24–1.97)||0.0 (0.8)|
|Americas||63||2.49 (2.23–2.77)||89.3 (<0.001)||5||1.12 (0.99–1.27)||0.0 (0.6)|
|Africa||8||2.01 (1.25–3.21)||88.5 (<0.001)||0||–||–|
|Renal cell cancer|
|Male||21||1.42 (1.25–1.62)||55.0 (0.001)||3||1.12 (0.85–1.47)||0.0 (0.6)|
|Female||19||1.32 (1.16–1.51)||26.6 (0.1)||–||–||–|
|Mixed||4||1.21 (1.10–1.33)||65.4 (<0.001)||5||1.23 (0.99–1.53)||70.3 (0.009)|
|Europe||16||1.02 (0.91–1.12)||0.0 (0.6)||1||1.30 (0.92–1.84)||–|
|Asia||6||1.19 (1.01–1.41)||27.9 (0.2)||3||1.12 (0.85–1.47)||0.0 (0.6)|
|Americas||40||1.35 (1.25–1.46)||57.5 (<0.001)||4||1.22 (0.95–1.58)||76.8 (0.005)|
|Oceania||2||1.74 (1.14–2.66)||70.2 (0.07)||0||–||–|
a Number of comparisons. Some studies include separate estimates for males and females and/or smoking category.
b p value for heterogeneity.
c The sum does not add up to the total number of studies in the meta–analysis since only studies reporting estimates separately for men and women were selected.
3.5. Secondhand smoking
The pooled RR of BC from secondhand smoking was 1.44 (95% CI 1.05–2.0; I2 = 59.8%, p = 0.021) and of RCC was 1.43 (95% CI 0.89–2.28; I2 = 55.3%, p = 0.08; data not shown). There were no data on DSM risk for SHS in this data set for either cancer type.
3.6.1. Tobacco products and bladder carcinogenesis
We found that tobacco consumption increases the risk of BC incidence and DSM, and we provide up-to-date and more precise quantitative estimates than previously available  . Although certain occupations (such dye workers) may have high individual risk elevations for BC, tobacco smoking appears to be responsible for most BC cases because of its high prevalence  .
Tobacco is a rich source of polycyclic aromatic hydrocarbons, aromatic amines, and N-nitroso compounds, which cause DNA damage via bulky adduct formation, single- and double-strand DNA breaks, and base modifications  . These acquired events complement an individual's genetic predisposition to smoking-related cancer. For example, first-degree relatives of BC patients have a 50–100% higher risk, which increases if the relative was diagnosed at <60 yr of age  and in a dose-dependent manner  .
Tobacco carcinogens are mostly metabolized by xenobiotic enzymes such as N-acetyltransferases (NATs) and glutathione S-transferases. These enzymes have alleles with different activity profiles. For example, individuals with slow NAT2 acetylation exhibit less efficient detoxification of carcinogens, leading to higher accumulation in urothelium. There is general consensus that individuals with slow NAT2 acetylation have a higher BC risk (up to 50%) and that this higher risk is mostly seen in smokers. Approximately 50% of individuals of European, 35% of African, and 15% of Asian descent may have slow acetlyation  .
Genome-wide association studies have recently focused on interactions between smoking and single nucleotide polymorphisms in BC patients, but a conclusive link has not been shown to date  .
3.6.2. Tobacco products and renal carcinogenesis
We found that RCC was 1.3-fold more common among smokers, in agreement with previous data  . In addition, the RCC DSM risk was 1.3-fold higher among current smokers. The triad of obesity, hypertension, and smoking are accepted as the main contributors to RCC , , and . It is thought that obesity confers risk through an increase in lipid peroxidation by-products that can cause DNA adducts  . It has also been shown that obese patients have higher circulating levels of insulin-like growth factor-1 (IGF-1) and vascular endothelial growth factor (VEGF), which have roles in cell proliferation. Patients with hypertension also have higher levels of lipid peroxidation by-products, and it is thought that hypertension results in renal tubular damage, making the kidney more susceptible to circulating carcinogens  and . It is thought that tobacco smoking adds to this and itself promotes the formation of oxygen free radicals that can cause DNA damage. Tobacco smoking leads to more aggressive RCC phenotypes, and patients who smoke at the time of nephrectomy have a lower survival rate  and . There is no universal consensus on whether this is due to direct effects of tobacco or the characteristics of smokers, who are perhaps less likely to seek health care and may suffer from delayed presentation.
A number of genes increase susceptibility to RCC, including von Hippel-Lindau (VHL)  . There are limited data on gene-environment interactions; however, in the last decade a link has been made between obesity and VHL tumor suppressor inactivation through mutations caused by reactive oxygen species  . Little is known about smoking and these interactions.
3.6.3. Patient outcomes
Smoking reduces perioperative performance status and impairs wound healing. Consequently, the risk of perioperative complications, disease progression, and tumor recurrence after treatment is higher  , as is the incidence of second smoking-related cancers after successful treatment  , among smokers when compared to nonsmokers. Disease-related patterns may differ between the malignancies. For RCC, smoking is associated with higher stage at diagnosis  . For BC, post-treatment recurrence risks were elevated in the majority of studies, although the hazard ratio (HR) varied in this review from 1.57 to 3.67 (data not shown)  and . Despite these outcomes, fewer than 50% of patients stop smoking after their cancer diagnosis  .
3.6.4. Secondhand smoking
One of the main methodologic limitations in measuring the effects of smoking on health outcomes is the difficulty in controlling for and measuring SHS exposure. In the articles included in our meta-analysis, researchers used household exposure, workplace exposure, or any environmental exposure methods to quantify SHS. However, these lack precision and make the strength of conclusions weaker than those for smokers.
There are various limitations to our study. In terms of search strategy and data collection, we chose to review only studies we found via the Medline database through PubMed, which may have limited the number of studies included. Furthermore, we only looked at studies written in English. However, a study by Moher et al  provides no evidence that language-restricted meta-analyses lead to biased estimates of intervention effectiveness.
In addition, there are concerns about the reliability and validity of smoking status questionnaires and interviews (smokers can under-report consumption or suffer recall bias). Most series were retrospective case-control studies, which may suffer from inaccurate documentation of smoking history. Prospective studies have fewer potential sources of bias, but under-reporting of smoking affects these studies too. Sweeney and Farrow  make the interesting point that smokers, who have poorer outcomes, may be under-represented because they deteriorate at an earlier stage compared to nonsmokers, and hence may not be available for studies. It is also accepted that SHS is hard to measure, and contamination is likely to confound risk estimates for nonsmokers  . It can be difficult to combine tobacco-smoking studies that may have looked at different tobacco-smoking combinations and used different definitions of smoking status. Hence, we chose not to analyze dose-response data (intensity of smoking) and instead used summary categories. Another potential pitfall of meta-analyses is the failure to appreciate the role of potentially confounding variables. To counter this, we used maximally adjusted risk estimates where provided. While we were not able to stratify for all characteristics (eg, ethnicity), we do report risk estimate differences by gender and geographic region. It would have been interesting to know whether the effect of tobacco smoking on BC incidence and DSM is similar in non–muscle-invasive (NMIBC) and muscle-invasive bladder cancer (MIBC), but studies included in this meta-analysis did not report results according to cancer stage. In 1987, Jensen et al found no difference in the effect of smoking on incidence between NMIBC and MIBC  .
Finally, during data analysis, Begg and Egger tests provided p values that were not significant for publication bias, even though visual inspection of funnel plots could not completely rule our this possibility.
We provide the largest meta-analysis to date on the relationship between tobacco smoking and BC and RCC incidence and mortality. Smoking involves a higher risk of cancer incidence and DSM, consistent with the literature. For BC, the incidence and DSM risk are greatest in current smokers and lowest in former smokers, indicating that cessation confers benefit. In 1988 smoking was responsible for 30–40% of BC and RCC cases  . Since then, some authors have suggested there has been an overall modest decrease in incidence and mainly mortality, particularly for BC , , and . Obesity is an increasing health problem and is probably partly responsible for the plateau in RCC incidence  and . Despite reductions in occupational exposures and smoking bans, smoking patterns in some countries remain high and the need to promote smoking cessation continues.
Author contributions: Marcus G. Cumberbatch had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Catto, Cumberbatch, La Vecchia.
Acquisition of data: Cumberbatch, Rota.
Analysis and interpretation of data: Cumberbatch, Rota.
Drafting of the manuscript: Cumberbatch, Rota, La Vecchia, Catto.
Critical revision of the manuscript for important intellectual content: Rota, La Vecchia.
Statistical analysis: Cumberbatch, Rota.
Obtaining funding: Cumberbatch, Catto.
Administrative, technical, or material support: Cumberbatch.
Supervision: La Vecchia.
Financial disclosures: Marcus G. Cumberbatch certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.
Funding/Support and role of the sponsor: None.
Acknowledgments: We would like to formally thank Dr. Francisco Real and Dr. Nuria Malats for their guidance and contribution to the structure and content of this work. Marcus G. Cumberbatch was supported by the Wellcome Trust and Matteo Rota was supported by a scholarship from the Italian Foundation for Cancer Research.
-  F. Islami, M. Stoklosa, J. Drope, A. Jemal. Global and regional patterns of tobacco smoking and tobacco control policies. Eur Urol Focus. 2015;1:3-16
-  C. Samanic, M. Kogevinas, M. Dosemeci, et al. Smoking and bladder cancer in Spain: effects of tobacco type, timing, environmental tobacco smoke, and gender. Cancer Epidemiol Biomarkers Prev. 2006;15:1348-1354 Crossref
-  Y.L. Zheng, S. Amr, D.A. Saleh, et al. Urinary bladder cancer risk factors in Egypt: a multicenter case-control study. Cancer Epidemiol Biomarkers Prev. 2012;21:537-546 Crossref
-  E. De Stefani, P. Correa, L. Fierro, E. Fontham, V. Chen, D. Zavala. Black tobacco, mate, and bladder cancer. A case-control study from Uruguay. Cancer. 1991;67:536-540 Crossref
-  I. Momas, J.P. Daures, B. Festy, J. Bontoux, F. Gremy. Bladder cancer and black tobacco cigarette smoking. Some results from a French case-control study. Eur J Epidemiol. 1994;10:599-604 Crossref
-  M. Burger, J.W. Catto, G. Dalbagni, et al. Epidemiology and risk factors of urothelial bladder cancer. Eur Urol. 2013;63:234-241 Crossref
-  M. Borzym-Kluczyk, I. Radziejewska, A. Zaniewska, et al. Effect of smoking on activity of N-acetyl-beta-hexosaminidase in serum and urine of renal cancer patients. Clin Biochem. 2009;42:1565-1567 Crossref
-  K. Rouissi, S. Ouerhani, B. Hamrita, et al. Smoking and polymorphisms in xenobiotic metabolism and DNA repair genes are additive risk factors affecting bladder cancer in Northern Tunisia. Pathol Oncol Res. 2011;17:879-886 Crossref
-  S. Chavan, F. Bray, J. Lortet-Tieulent, M. Goodman, A. Jemal. International variations in bladder cancer incidence and mortality. Eur Urol. 2014;66:59-73 Crossref
-  A. Znaor, J. Lortet-Tieulent, M. Laversanne, A. Jemal, F. Bray. International variations and trends in renal cell carcinoma incidence and mortality. Eur Urol. 2015;67:519-530 Crossref
-  J. Benichou, W.H. Chow, J.K. McLaughlin, J.S. Mandel, J.F. Fraumeni Jr. Population attributable risk of renal cell cancer in Minnesota. Am J Epidemiol. 1998;148:424-430 Crossref
-  H. Chu, M. Wang, Z. Zhang. Bladder cancer epidemiology and genetic susceptibility. J Biomed Res. 2013;27:170-178 Crossref
-  R.J. Lammers, W.P. Witjes, K. Hendricksen, C.T. Caris, M.H. Janzing-Pastors, J.A. Witjes. Smoking status is a risk factor for recurrence after transurethral resection of non–muscle-invasive bladder cancer. Eur Urol. 2011;60:713-720 Crossref
-  C. Lee, K.H. Kim, D. You, et al. Smoking and survival after radical cystectomy for bladder cancer. Urology. 2012;80:1307-1312 Crossref
-  J.P. Sfakianos, S.F. Shariat, R.L. Favaretto, J. Rioja, H.W. Herr. Impact of smoking on outcomes after intravesical bacillus Calmette-Guerin therapy for urothelial carcinoma not invading muscle of the bladder. BJU Int. 2011;108:526-530 Crossref
-  M. Maclure, R.B. Katz, M.S. Bryant, P.L. Skipper, S.R. Tannenbaum. Elevated blood levels of carcinogens in passive smokers. Am J Public Health. 1989;79:1381-1384 Crossref
-  D. Moher, A. Liberati, J. Tetzlaff, D.G. Altman, P. Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097 Crossref
-  S. Greenland. Quantitative methods in the review of epidemiologic literature. Epidemiol Rev. 1987;9:1-30
-  R. DerSimonian, N. Laird. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177-188 Crossref
-  J.P. Higgins, S.G. Thompson. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539-1558 Crossref
-  M. Egger, G. Davey Smith, M. Schneider, C. Minder. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629-634 Crossref
-  C.B. Begg, M. Mazumdar. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50:1088-1101 Crossref
-  P.M. Marcus, R.B. Hayes, P. Vineis, et al. Cigarette smoking, N-acetyltransferase 2 acetylation status, and bladder cancer risk: a case-series meta-analysis of a gene-environment interaction. Cancer Epidemiol Biomarkers Prev. 2000;9:461-467
-  M.C. Stern, J. Lin, J.D. Figueroa, et al. Polymorphisms in DNA repair genes, smoking, and bladder cancer risk: findings from the international consortium of bladder cancer. Cancer Res. 2009;69:6857-6864 Crossref
-  C. Pelucchi, C. Bosetti, E. Negri, M. Malvezzi, C. La Vecchia. Mechanisms of disease: The epidemiology of bladder cancer. Nat Clin Pract Urol. 2006;3:327-340
-  U. Gabriel, L. Li, C. Bolenz, et al. New insights into the influence of cigarette smoking on urothelial carcinogenesis: smoking-induced gene expression in tumor-free urothelium might discriminate muscle-invasive from nonmuscle-invasive urothelial bladder cancer. Mol Carcinog. 2012;51:907-915 Crossref
-  N. Rothman, M. Garcia-Closas, N. Chatterjee, et al. A multi-stage genome-wide association study of bladder cancer identifies multiple susceptibility loci. Nat Genet. 2010;42:978-984 Crossref
-  J.D. Hunt, O.L. van der Hel, G.P. McMillan, P. Boffetta, P. Brennan. Renal cell carcinoma in relation to cigarette smoking: meta-analysis of 24 studies. Int J Cancer. 2005;114:101-108 Crossref
-  G. Corrao, L. Scotti, V. Bagnardi, R. Sega. Hypertension, antihypertensive therapy and renal-cell cancer: a meta-analysis. Curr Drug Saf. 2007;2:125-133 Crossref
-  M.P. Purdue, L.E. Moore, M.J. Merino, et al. An investigation of risk factors for renal cell carcinoma by histologic subtype in two case-control studies. Int J Cancer. 2013;132:2640-2647 Crossref
-  World Cancer Research Fund/American Institute for Cancer Research. Food, nutrition, physical activity, and the prevention of cancer: a global perspective. (AICR, Washington, DC, 2007)
-  J.K. McLaughlin, L. Lipworth, R.E. Tarone. Epidemiologic aspects of renal cell carcinoma. Semin Oncol. 2006;33:527-533 Crossref
-  A. Parker, C. Lohse, J. Cheville, B. Leibovich, T. Igel, M. Blute. Evaluation of the association of current cigarette smoking and outcome for patients with clear cell renal cell carcinoma. Int J Urol. 2008;15:304-308 Crossref
-  V.W. Setiawan, D.O. Stram, A.M. Nomura, L.N. Kolonel, B.E. Henderson. Risk factors for renal cell cancer: the multiethnic cohort. Am J Epidemiol. 2007;166:932-940 Crossref
-  J.M. Underwood, J.S. Townsend, E. Tai, A. White, S.P. Davis, T.L. Fairley. Persistent cigarette smoking and other tobacco use after a tobacco-related cancer diagnosis. J Cancer Surviv. 2012;6:333-344 Crossref
-  M.S. Shiels, T. Gibson, J. Sampson, et al. Cigarette smoking prior to first cancer and risk of second smoking-associated cancers among survivors of bladder, kidney, head and neck, and stage I lung cancers. J Clin Oncol. 2014;32:3989-3995 Crossref
-  E.C. Kauffman, C.J. Ricketts, S. Rais-Bahrami, et al. Molecular genetics and cellular features of TFE3 and TFEB fusion kidney cancers. Nat Rev Urol. 2014;11:465-475 Crossref
-  R.P. Theis, S.M. Dolwick Grieb, D. Burr, T. Siddiqui, N.R. Asal. Smoking, environmental tobacco smoke, and risk of renal cell cancer: a population-based case-control study. BMC Cancer. 2008;8:387 Crossref
-  D. Moher, B. Pham, T.P. Klassen, et al. What contributions do languages other than English make on the results of meta-analyses?. J Clin Epidemiol. 2000;53:964-972 Crossref
-  C. Sweeney, D.C. Farrow. Differential survival related to smoking among patients with renal cell carcinoma. Epidemiology. 2000;11:344-346 Crossref
-  J. Siemiatycki, R. Dewar, D. Krewski, M. Desy, L. Richardson, E. Franco. Are the apparent effects of cigarette smoking on lung and bladder cancers due to uncontrolled confounding by occupational exposures?. Epidemiology. 1994;5:57-65 Crossref
-  V.L. Ernster. Trends in smoking, cancer risk, and cigarette promotion. Current priorities for reducing tobacco exposure. Cancer. 1988;62:1702-1712 Crossref
-  C. Bosetti, P. Bertuccio, M. Malvezzi, et al. Cancer mortality in Europe, 2005–2009, and an overview of trends since 1980. Ann Oncol. 2013;24:2657-2671 Crossref
-  F. Levi, J. Ferlay, C. Galeone, et al. The changing pattern of kidney cancer incidence and mortality in Europe. BJU Int. 2008;101:949-958 Crossref
-  S.A. Strope, J.E. Montie. The causal role of cigarette smoking in bladder cancer initiation and progression, and the role of urologists in smoking cessation. J Urol. 2008;180:31-37 Crossref
a Academic Urology Unit, University of Sheffield, The Medical School, Beech Hill Road, Sheffield, UK
b Department of Epidemiology, IRCCS Istituto di Ricerche Farmacologiche Mario Negri, Milan, Italy
c Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy
⁎ Corresponding author. Academic Urology Unit, University of Sheffield, G Floor, The Medical School, Beech Hill Road, Sheffield S10 2RX, UK. Tel. +44 114 2261229; Fax: +44 114 2712268.
© 2015 European Association of Urology, Published by Elsevier B.V.