Assessment of the Ecological Association between Tobacco Smoking Exposure and Bladder Cancer Incidence over the Past Half-Century in the United States

Round 1

Reviewer 1 Report

See comments in attached pdf.

Smoking is associated with a 15 to 30 times risk of lung cancer, but only a 3 times risk of bladder cancer. Couldn't a 10-fold difference in risks explain your results?

https://www.cdc.gov/cancer/lung/basic_info/risk_factors.htm#:~:text=People%20who%20smoke%20cigarettes%20are,the%20risk%20of%20lung%20cancer.

https://health.clevelandclinic.org/surprising-link-smoking-and-bladder-cancer-2/

 

Comments for author File: Comments.pdf

Author Response

Reviewer 1:

Smoking is associated with a 15 to 30 times risk of lung cancer, but only a 3 times risk of bladder cancer. Couldn't a 10-fold difference in risks explain your results?

https://www.cdc.gov/cancer/lung/basic_info/risk_factors.htm#:~:text=People%20who%20smoke%20cigarettes%20are,the%20risk%20of%20lung%20cancer.

https://health.clevelandclinic.org/surprising-link-smoking-and-bladder-cancer-2/

We thank the reviewer for this important comment. Theoretically, a higher relative risk (odds ratio) between smoking and lung versus bladder cancer might affect the incidence rate difference (IRD), the absolute difference in the risk of cancer between smokers and non-smokers. Thus, a higher relative risk for cancer between smokers and non-smokers might also be associated with a higher absolute risk.

One main advantage of ecological studies is that it enables us to observe the pattern in trends between smoking and the incidence of bladder versus lung cancer. However, due to the ecological nature of this study, individual information cannot be retrieved. Therefore, we cannot know whether those who smoked got cancer.

We have added the following sentences and citations in the discussion and limitation sections to acknowledge this:

Discussion on lines 289-294:

“Additionally, there is evidence that the relative risk of smoking is higher for lung than for bladder cancer, which might impact the IRD, a measure of the absolute difference in the risk of cancer between smokers and non-smokers [29,30]. This might explain why our observations highlight a lack of congruence between tobacco smoking and incidence of bladder compared to that of lung cancer. Nevertheless, the takeaway message is that other risk factors may be at play for bladder cancer.”    

29. Division of Cancer Prevention and Control, Centers for Disease Control and Prevention. What Are the Risk Factors for Lung Cancer? Available online: https://www.cdc.gov/cancer/lung/basic_info/risk_factors.htm#:~:text=People%20who%20smoke%20cigarettes%20are,the%20risk%20of%20lung%20cancer (accessed on January 18, 2023).
30. ‘Surprising Link’: Smoking and Bladder Cancer. Available online: https://health.clevelandclinic.org/surprising-link-smoking-and-bladder-cancer-2/ (accessed on January 18, 2023).

Limitation on lines 306-309:

“Our results are not devoid of limitations. First, since this is an ecological study, data on individual people are not available (including adjusting for pack-years smoked), and therefore we do not know if the smokers captured between 1953 and 1983 in this study developed bladder or lung cancer.”

Reviewer 2 Report

Dear Authors,

-Your references are not updated, so please add more updated (last 5 years) references to cover up current status of the issue.

-Your selection criterias are not clear. If it is possible, give them in a table.

-I suggest you to use another statistical model to confirm your results. Only one test is not sufficient to emphasize overall meaning.

You need to revise these:

-Figure 1 is very complicated and small. It is so hard to read especially years and axis'. Maybe you should give the results in one diagram with different colors or seperate the diagrams and add them as the supplementary part.

-Figure 2 has low resolution, therefore it is not easy to read. Also error bars can not be seen.

-Figure 3 has no error bars.

Discussion and conclusion should be more comprehensive. You need to discuss your results with the existing literature. Also in the conclusion part, you need to be more clear to understand the meaning of the study. Please highlight your study in terms of importance in the cancer research field. 

 

Author Response

Reviewer 2:

Dear Authors,

-Your references are not updated, so please add more updated (last 5 years) references to cover up current status of the issue.

Thank you very much for this important comment. We have updated several of our references to cover the contemporary literature on tobacco smoking and bladder cancer. The following references are replacing older studies:

5. Ng, K.L. The Etiology of Bladder Cancer. In Urologic Cancers, Barber, N., Ali, A., Eds.; Exon Publications

Copyright: The Authors.; The author confirms that the materials included in this chapter do not violate copyright laws. Where relevant, appropriate permissions have been obtained from the original copyright holder(s), and all original sources have been appropriately acknowledged or referenced.: Brisbane (AU), 2022.

10. Zhao, X.; Wang, Y.; Liang, C. Cigarette smoking and risk of bladder cancer: a dose-response meta-analysis. Int Urol Nephrol 2022, 54, 1169-1185, doi:10.1007/s11255-022-03173-w.
11. van Osch, F.H.M.; Vlaanderen, J.; Jochems, S.H.J.; Bosetti, C.; Polesel, J.; Porru, S.; Carta, A.; Golka, K.; Jiang, X.; Stern, M.C.; et al. Modeling the Complex Exposure History of Smoking in Predicting Bladder Cancer: A Pooled Analysis of 15 Case-Control Studies. Epidemiology 2019, 30, 458-465, doi:10.1097/ede.0000000000000964.
12. Masaoka, H.; Matsuo, K.; Oze, I.; Kimura, T.; Tamakoshi, A.; Sugawara, Y.; Tsuji, I.; Sawada, N.; Tsugane, S.; Ito, H.; et al. Cigarette smoking, smoking cessation and bladder cancer risk: A pooled analysis of ten cohort studies in Japan. J Epidemiol 2022, doi:10.2188/jea.JE20220085.
13. Lobo, N.; Afferi, L.; Moschini, M.; Mostafid, H.; Porten, S.; Psutka, S.P.; Gupta, S.; Smith, A.B.; Williams, S.B.; Lotan, Y. Epidemiology, Screening, and Prevention of Bladder Cancer. Eur Urol Oncol 2022, 5, 628-639, doi:10.1016/j.euo.2022.10.003.
14. Bellamri, M.; Walmsley, S.J.; Brown, C.; Brandt, K.; Konorev, D.; Day, A.; Wu, C.F.; Wu, M.T.; Turesky, R.J. DNA Damage and Oxidative Stress of Tobacco Smoke Condensate in Human Bladder Epithelial Cells. Chem Res Toxicol 2022, 35, 1863-1880, doi:10.1021/acs.chemrestox.2c00153.
15. Teoh, J.Y.; Huang, J.; Ko, W.Y.; Lok, V.; Choi, P.; Ng, C.F.; Sengupta, S.; Mostafid, H.; Kamat, A.M.; Black, P.C.; et al. Global Trends of Bladder Cancer Incidence and Mortality, and Their Associations with Tobacco Use and Gross Domestic Product Per Capita. Eur Urol 2020, 78, 893-906, doi:10.1016/j.eururo.2020.09.006.

-Your selection criterias are not clear. If it is possible, give them in a table.

This is a valid point. We have rephrased the following in the methods section on lines 75-84:

“Our primary independent variable was the annual prevalence of tobacco smoking between 1953 and 1983. Current smoker were defined as individuals ≥18 years old who reported smoking ≥ 100 cigarettes during their lifetime, and who, at the time they participated in the National Health Interview Survey, reported smoking every day or some days [18]. Since there is no single data source covering 1953 to 1983, we relied on both the Report of the Surgeon General [19], the Center for Disease Control [20], to derive the overall and gender specific annual prevalence of tobacco smoking in the United States over this period.

18. Centers for Disease Control and Prevention: National Center for Health Statistics (US). National Health Interview Survey. Available online: https://www.cdc.gov/nchs/nhis/data-questionnaires-documentation.htm (accessed on January 18 2023).
19. Pinney, J.M. United States. Public Health Service. Office of the Assistant Secretary for Health. Office on Smoking and Health. A Report of the Surgeon General: Appendix: Cigarette Smoking in the United States, 1950-1978 (pages A-1 through A-29); 1979.
20. Centers for Disease Control and Prevention. Morbidity and Mortality Weekly Reports (MMWR). Available online: https://www.cdc.gov/tobacco/data_statistics/mmwrs/index.htm (accessed on January 19, 2023)

-I suggest you to use another statistical model to confirm your results. Only one test is not sufficient to emphasize overall meaning.

We appreciate the reviewer’s feedback. We have developed the following statistical plan under direct supervision and feedback from our senior Biostatistician:

To estimate the incidence rate differences, we modeled separate linear regressions of the tobacco smoking prevalence and each malignancy (bladder and lung). The regression coefficients – which represent the incidence rate differences - were estimated using weighted least squares, with weights equal to the inverse of the estimated variance of the yearly incidence rate for each malignancy.

Then, to compare whether the prevalence of tobacco smoking differentially predicts bladder versus lung cancer we have conducted the following two statistical tests:

1. A Wald test
2. R-squared statistic derived from the weighted least squares regression models (described above).

If the reviewer has any suggestions, we would be pleased to provide additional analyses.

You need to revise these:

-Figure 1 is very complicated and small. It is so hard to read especially years and axis'. Maybe you should give the results in one diagram with different colors or seperate the diagrams and add them as the supplementary part.

-Figure 2 has low resolution, therefore it is not easy to read. Also error bars can not be seen.

We thank the reviewer for their comment. We agree with the reviewer. Accordingly, we have made the figures larger.

Please note that error bars for Figure 2 are included; however, they might be more difficult to notice in some of the graphs due to the very narrow confidence interval (in men >> women). Thank you for your understanding.

-Figure 3 has no error bars.

Thank you for your valuable feedback. Figure 3 is only a graphic representation of the R-squared statistic derived from the weighted least squares regression model to estimate the proportion of smoking-related changes in the incidence of each malignancy in the overall, male, and female populations. Since there are no confidence intervals associated with R-squared statistic, no error bars were included in the graphical illustration. Please refer below to the software output for lung cancer in the overall population as an example.

  / * Please refer to the attached containing the output*/

Discussion and conclusion should be more comprehensive. You need to discuss your results with the existing literature. Also in the conclusion part, you need to be more clear to understand the meaning of the study. Please highlight your study in terms of importance in the cancer research field. 

Thank you very much for this important feedback. We have made the following changes to further discuss our results within the existing literature and to highlight the study in terms of importance in the cancer research field:

Discussion on lines 289-294:

“Additionally, there is evidence that the relative risk of smoking is higher for lung than for bladder cancer, which might impact the IRD, a measure of the absolute difference in the risk of cancer between smokers and non-smokers [33,34]. This might explain why our observations highlight a lack of congruence between tobacco smoking and incidence of bladder compared to that of lung cancer. Nevertheless, the takeaway message is that other risk factors may be at play for bladder cancer.”  

33. Division of Cancer Prevention and Control, Centers for Disease Control and Prevention. What Are the Risk Factors for Lung Cancer? Available online: https://www.cdc.gov/cancer/lung/basic_info/risk_factors.htm#:~:text=People%20who%20smoke%20cigarettes%20are,the%20risk%20of%20lung%20cancer (accessed on January 18, 2023).
34. ‘Surprising Link’: Smoking and Bladder Cancer. Available online: https://health.clevelandclinic.org/surprising-link-smoking-and-bladder-cancer-2/ (accessed on January 18, 2023).

Discussion on lines 257-264:

“Teoh et al. used the World Health Organization Global Health Observatory database to examine gender-specific incidence and mortality trends of bladder cancer. They found that tobacco use was weakly associated with bladder cancer incidence among men (r=0.20), but strongly associated with bladder cancer incidence among women (r=0.67) [29]. Thus, occupational exposures – more prominent among men than women - might explain why despite a decreasing trend in prevalence of tobacco smoking among men, the incidence rate for bladder cancer is stable.”

29. Teoh, J.Y.; Huang, J.; Ko, W.Y.; Lok, V.; Choi, P.; Ng, C.F.; Sengupta, S.; Mostafid, H.; Kamat, A.M.; Black, P.C.; et al. Global Trends of Bladder Cancer Incidence and Mortality, and Their Associations with Tobacco Use and Gross Domestic Product Per Capita. Eur Urol 2020, 78, 893-906, doi:10.1016/j.eururo.2020.09.006.

 

Conclusion on lines 336-344:

“While smoking is undeniably associated with bladder and lung cancer, our findings suggest that other risk factors may contribute to the incidence of bladder cancer, in addition to tobacco smoking. This highlights the need to further investigate the potential impact of environmental and occupational exposures, as well as genetic predispositions, on bladder cancer risk. Additionally, the study provides evidence that increased awareness of bladder cancer risk factors is necessary in order to reduce bladder cancer incidence and mortality rates.”

Author Response File: Author Response.docx

Reviewer 3 Report

This is a well-prepared ecological study on smoking and bladder cancer. 

The study is interesting and its findings are important.

Please consider some minor revisions:

Abstract: lines 22-25 may be replaced with data on their own findings. Statistical methods may be described in the main text. However, this is just a suggestion and the Authors should decide.

Methods: lines 78-83 - please remove links to websites and replace them with references. E.g., the CDC website should be replaced with a reference.

Discussion: please provide 2-3 sentences on practical implications of this study.

 

Author Response

 Reviewer 3:

This is a well-prepared ecological study on smoking and bladder cancer. 

The study is interesting and its findings are important.

Please consider some minor revisions:

Abstract: lines 22-25 may be replaced with data on their own findings. Statistical methods may be described in the main text. However, this is just a suggestion and the Authors should decide.

Thank you very much for your thorough review. We appreciate your comment. The methods section in the abstract is mandatory as per Current Oncology. We understand that the word limit of the abstract might prevent us from including all our findings and we apologize for this inconvenience.

Methods: lines 78-83 - please remove links to websites and replace them with references. E.g., the CDC website should be replaced with a reference.

Thank you for the suggestion. The links to the websites were replaces with references in the methods section on lines 77-83.

Discussion: please provide 2-3 sentences on practical implications of this study.

Thank you once again. We have added the following on lines 336-343:

“While smoking is undeniably associated with bladder and lung cancer, our findings suggest that other risk factors may contribute to the incidence of bladder cancer, in addition to tobacco smoking. This highlights the need to further investigate the potential impact of environmental and occupational exposures, as well as genetic predispositions, on bladder cancer risk. Additionally, the study provides evidence that increased awareness of bladder cancer risk factors is necessary in order to reduce bladder cancer incidence and mortality rates.”

Reviewer 4 Report

Estimated Authors,

I've been invited to review this paper from Seisen et al. on the Association Between Tobacco Smoking Exposure and Bladder Cancer Incidence Over the Past Half-Century in the United States. Through an ecological design, Authors have assessed whether Bladder Cancer, compared to Lung cancer, may have been influenced by smoking habits from the general population. In this analysis, results are both consistent with several previous studies (and Authors have accurately addressed these results in their discussion section), and surprising, as these findings seemingly demonstrate that the trends in tobacco smoking explain most of the trends in incidence rates for lung cancer, but not for bladder cancer. Several explanations could be advocated, being all quite unsatisfying and will require further studies to be detailed. Being an ecological study, i.e. an hypothesis-generating studies, this research has, from my point of view, substantially fulfilled all the potential aims, but some improvements are collectively required before its eventual acceptance.

First of all, some further explanation about SEER registries should be provided: their source of data, their quality assessment. Moreover, Authors should clarify that the overall coverage of US population is seemingly incomplete (https://seer.cancer.gov/registries/list.html), with a potential over-representation of native-american people that should be at least discussed and eventually ruled out (in case). 

Another shortcoming is that neither lung cancer nor bladder cancer are particularly specific terms. Bladder cancer encompasses a quite diverse series of neoplasia, i.e. squamous cell carcinoma (1-2% of all bladder cancers in the US), adenocarcinoma (1%), small cell carcinoma (1%), urothelial carcinoma (95%); moreover, even within urothelial carcinoma, invasive and non invasive subtypes could be identified. Similarly, lung cancer includes squamous cells and adenocarcinoma, both of them characterized by distinctive association with smoking history. In other words, all of these potential shortcomings should be discussed among the quite honest limit section that Authors have included (and whose appropriateness I've appreciated).

 

Author Response

 Reviewer 4:

Estimated Authors,

I've been invited to review this paper from Seisen et al. on the Association Between Tobacco Smoking Exposure and Bladder Cancer Incidence Over the Past Half-Century in the United States. Through an ecological design, Authors have assessed whether Bladder Cancer, compared to Lung cancer, may have been influenced by smoking habits from the general population. In this analysis, results are both consistent with several previous studies (and Authors have accurately addressed these results in their discussion section), and surprising, as these findings seemingly demonstrate that the trends in tobacco smoking explain most of the trends in incidence rates for lung cancer, but not for bladder cancer. Several explanations could be advocated, being all quite unsatisfying and will require further studies to be detailed. Being an ecological study, i.e. an hypothesis-generating studies, this research has, from my point of view, substantially fulfilled all the potential aims, but some improvements are collectively required before its eventual acceptance.

 

Thank you very much for your feedback.

First of all, some further explanation about SEER registries should be provided: their source of data, their quality assessment. Moreover, Authors should clarify that the overall coverage of US population is seemingly incomplete (https://seer.cancer.gov/registries/list.html), with a potential over-representation of native-american people that should be at least discussed and eventually ruled out (in case). 

Thank you for your important comment. We have made the following changes in the Methods and the Limitations sections to incorporate your feedback.

Methods on lines 86-95:

“We used the Surveillance, Epidemiology, and End Results (SEER)–9 registries, which provides a population-based sample representing around 28% of the US population [21], to obtain age-adjusted incidence rates for bladder cancer by identifying adults ≥ 40 years old with a histologically confirmed diagnosis of primary bladder cancer (ICD-O-3 site codes: C67.0-C67.9). As a comparator, we obtained the age-adjusted incidence rates for primary lung cancer between 1983 and 2013 by identifying adults aged ≥ 40 years old with a histologically confirmed diagnosis of lung cancer (ICD-O-3 site codes: C34.0-C34.9). Since SEER is a population-based dataset with cancer-specific outcomes, it is a reliable resource for epidemiological studies [21].”

21. Cole, A.P.; Friedlander, D.F.; Trinh, Q.D. Secondary data sources for health services research in urologic oncology. Urol Oncol 2018, 36, 165-173, doi:10.1016/j.urolonc.2017.08.008.

Limitation on lines 311-313:

“Third, although SEER is a population-based dataset that is representative of the US population, it might oversample for Non-Hispanic American Indian/Alaska Native [41].”

41. National Cancer Institute. Surveillance, Epidemiology, and End Results Program. Race and Hispanic Ethnicity Changes. Available online: https://seer.cancer.gov/seerstat/variables/seer/race_ethnicity/ (accessed on January 24, 2023).

Another shortcoming is that neither lung cancer nor bladder cancer are particularly specific terms. Bladder cancer encompasses a quite diverse series of neoplasia, i.e. squamous cell carcinoma (1-2% of all bladder cancers in the US), adenocarcinoma (1%), small cell carcinoma (1%), urothelial carcinoma (95%); moreover, even within urothelial carcinoma, invasive and non invasive subtypes could be identified. Similarly, lung cancer includes squamous cells and adenocarcinoma, both of them characterized by distinctive association with smoking history. In other words, all of these potential shortcomings should be discussed among the quite honest limit section that Authors have included (and whose appropriateness I've appreciated).

Thank you very much for raising this important point. We totally agree that smoking might differentially explain the pathology of the disease. Accorindgly, we acknowledged this in our limitation section on lines 309-311:

“Second, we have not accounted for the pathology of bladder and lung cancer as tobacco smoking might variably account for the type of pathology.”

Reviewer 5 Report

It is an interesting epidemiological study

Please clarify the mechanisms by which tobacco acts on the bladder.

It is important to stress the concept that smoking duration affects the onset of the disease.

Why did you associate age with current smoking?

A paragraph about statistical methods should be included

Do you have data about e-cigarette induced lung cancer?

More details should be provided concerning pack-years and lung cancer prevalent staging.

Please add the following references

-Future Sci OA. 2019 May 3;5(5):FSO394.

-Future Oncol. 2013 May;9(5):649-55.

Author Response

 Reviewer 5:

It is an interesting epidemiological study

Please clarify the mechanisms by which tobacco acts on the bladder.

It is important to stress the concept that smoking duration affects the onset of the disease.

Thank you very much for your thorough review. We have added the following to the introduction to clarify the mechanism smoking may lead to bladder cancer and to stress on the concept that smoking duration affects the onset of the disease:

Lines 49-54:

“There is evidence that exposure to aromatic amines through smoking may cause DNA adduction and mutagenesis [7-9]. In recent meta-analyses, the risk of developing bladder cancer was found to be higher among current smokers compared to former smokers [10,11]. Moreover, reports showed that, for an equal total exposure in pack-years smoked, smoking less for a longer duration is more harmful than smoking more for a shorter duration [11].

7. Besaratinia, A.; Tommasi, S. Genotoxicity of tobacco smoke-derived aromatic amines and bladder cancer: current state of knowledge and future research directions. The FASEB Journal 2013, 27, 2090-2100, doi:https://doi.org/10.1096/fj.12-227074.
8. Pezzuto, A.; Citarella, F.; Croghan, I.; Tonini, G. The effects of cigarette smoking extracts on cell cycle and tumor spread: novel evidence. Future Sci OA 2019, 5, Fso394, doi:10.2144/fsoa-2019-0017.
9. Tonini, G.; D'Onofrio, L.; Dell'Aquila, E.; Pezzuto, A. New molecular insights in tobacco-induced lung cancer. Future Oncol 2013, 9, 649-655, doi:10.2217/fon.13.32.
10. Zhao, X.; Wang, Y.; Liang, C. Cigarette smoking and risk of bladder cancer: a dose-response meta-analysis. Int Urol Nephrol 2022, 54, 1169-1185, doi:10.1007/s11255-022-03173-w.
11. van Osch, F.H.M.; Vlaanderen, J.; Jochems, S.H.J.; Bosetti, C.; Polesel, J.; Porru, S.; Carta, A.; Golka, K.; Jiang, X.; Stern, M.C.; et al. Modeling the Complex Exposure History of Smoking in Predicting Bladder Cancer: A Pooled Analysis of 15 Case-Control Studies. Epidemiology 2019, 30, 458-465, doi:10.1097/ede.0000000000000964.

Why did you associate age with current smoking?

Thank you very much for this important question. Age-adjusted incidence rates are used to compare rates of cancer between different populations, or between the same population over different time periods, in order to account for differences in age distribution. Incidence rates are calculated by dividing the number of new cases over a given period of time by the size of the population at risk during that period. However, age is a major factor in determining risk for certain types of cancers. Therefore, age-adjusted incidence rates are used to take into account the differences in age distribution between populations in order to make meaningful comparisons.

A paragraph about statistical methods should be included

Thank you. We have detailed our analyses in the section titled “Statistical Analyses” between lines 96 and 122. We have provided detailed steps to allow the reproducibility of our results. We are happy to provide further information or make any amendments if you think it might not be clear to the readers.

Do you have data about e-cigarette induced lung cancer?

Thank you for your question. The objective of the study is to compare the trend in incidence of tobacco smoking to trend in bladder and lung cancer incidences with a 30-year latency period. While it would be a very interesting question, we are looking at the prevalence of tobacco smoking between 1953 and 1983 and examining the cancer incidence between 1983 and 2013. Unfortunately, the e-cigarette was not commercialized until 2006 and it might require another decade to further understand its impact on cancer incidence.

More details should be provided concerning pack-years and lung cancer prevalent staging.

Thank you for your feedback. Since this is an ecological study attempting to understand how the exposure to tobacco smoking and cancer incidence, it would not possible to obtain individual information. Accordingly, we have acknowledged that in our limitation section on lines 306-309:

“Our results are not devoid of limitations. First, since this is an ecological study, data on individual people are not available (including adjusting for pack-years smoked), and therefore we do not know if the smokers captured between 1953 and 1983 in this study developed bladder or lung cancer.”

Please add the following references

-Future Sci OA. 2019 May 3;5(5):FSO394.

-Future Oncol. 2013 May;9(5):649-55.

Thank you for the suggestion. They were added on line 50.

8. Pezzuto, A.; Citarella, F.; Croghan, I.; Tonini, G. The effects of cigarette smoking extracts on cell cycle and tumor spread: novel evidence. Future Sci OA 2019, 5, Fso394, doi:10.2144/fsoa-2019-0017.
9. Tonini, G.; D'Onofrio, L.; Dell'Aquila, E.; Pezzuto, A. New molecular insights in tobacco-induced lung cancer. Future Oncol 2013, 9, 649-655, doi:10.2217/fon.13.32.

Round 2

Reviewer 1 Report

The authors have addressed my concerns by stating the difference in relative risk of bladder cancer and lung cancer associated with smoking, and how this difference might affect the outcome of the study.

Reviewer 2 Report

Thanks for your revisions.