Racial differences in Urinary Bladder Cancer in the United States

Urinary bladder cancer (UBC) has a high incidence rates in many southern and eastern European countries, in parts of Africa and the Middle East, and in North America. It exhibits a wide variety of histological types that goes from less aggressive to rapid-growing ones. In order to compare the different presentations, etiologies, and prognoses among racial groups, including NHW (non-Hispanic white), HW (Hispanic white), blacks, and API (Asian and Pacific Islander), we analyzed the UBC patients diagnosed between 1973 and 2014 using SEER (Surveillance, Epidemiology, and End Results) database. Patient characteristics, age-adjusted incidence rates, and survival were compared across races. There are significant racial differences in patients’ characteristics, including gender, marital status, age at diagnosis, treatment strategies, grade, stage, survival time, and so on. Overall, non-Hispanic whites have the highest incidence rate, followed by blacks, Hispanic whites, and APIs. In the analysis of survival, significant racial differences exist when stratified by gender, age group, histological type, stage, location and treatment strategies. Racial differences exist among UBC patients in the United States in terms of characteristics, incidence, and survival. Future studies may collect and analyze more data for comprehensive description and interpretation of the racial differences.

SCientifiC RepoRts | (2018) 8:12521 | DOI: 10.1038/s41598-018-29987-2 of cancer, the grade, and the stage and so on. The most commonly used treatment strategy is transurethral resection of the bladder tumor. Other treatments include the cystectomy, the intravesical chemotherapy, the systemic chemotherapy, the radiation therapy, and the immunotherapy 13 . Cystectomy has the best treatment effect. It can prevent recurrence to a large extent. Transurethral resection of the bladder tumor followed by combined chemotherapy and radiation therapy also can achieve as effective as cystectomy 14 . Although the five-year survival rate of UBC is good (62-88%), the recurrence rate is also high 15 . There are a lot of prognostic factors for recurrence of UBC, including the number of tumors present at diagnosis, the recurrence rate in the previous period, the size of tumor, and the grade of tumor 16,17 .
This study aims to comprehensively describe the racial differences among UBC patients in the United States in terms of patients' characteristics, clinical-pathologic features, incidence rates, and survival rates by analyzing the SEER database. Recently, researches on racial differences among cancer patients have attracted extensive attention [1][2][3] , but the discussion on UBC remains scarce. The available UBC studies may have limitations by either the numbered racial groups or the specified outcomes. For instance, Scosyrev and others 18 only compared blacks and whites. They found that African American patients present with less favorable tumor characteristics compared with white patients, accounting for a much higher UBC mortality experienced by these groups. Burger and others 10 (2013) mainly studied on risk factors of urothelial bladder cancer. Although the racial difference of UBC are mentioned in his article, the discussion only focuses on survival outcome as well as the discussion about racial difference of UBC is not deep enough. Yee and others 19 also just focus on survival outcome for different racial groups in UBC. They found that ethnic disparities in bladder cancer survival persist between whites and blacks, whereas survival in other ethnic minority groups appears similar to that of whites. However, our study on UBC are more comprehensive than previous study. We compare a wider spectrum of racial groups, including NHW (non-Hispanic white), HW (Hispanic white), blacks, and API (Asian and Pacific Islander), and analyze multiple aspects, including patients' characteristics, clinical-pathological features, incidence rate, and survival rate.

Methods
Source population. The Surveillance, Epidemiology, and End Results (SEER, http://seer.cancer.gov) database of the National Cancer Institute is utilized to provide population-based sample data in this study. This program collects information on cancer statistics in an effort to reduce the cancer burden among the U.S. population. Approximately 9.5%, 14%, and 28% of the U.S. population are analyzed by covering registry 9, 13, and 18 respectively, and the International Classification of Disease for Oncology (ICD-O-3) site code C670-C679 are referred to identify UBC cases in those population. Tumor sites and histological types are coded based on WHO specified criteria in ICD-O-3. And histological types are also grouped by ICD-O-3. Patients with PTCC, NPTCC, Carcinoma, and SCC are identified by the histology codes 8130-8131, 8120-8124, 8010-8015, and 8070-8078, respectively. The division of treatment variable is classified comprehensively in this study. There are three main treatment variables in the SEER database, including surgery, radiotherapy, and chemotherapy. We consider a variety of combinations of these three methods including surgery, chemotherapy, radiation, SR (surgery and radiation), SC (surgery and chemotherapy), RC (radiation and chemotherapy), SRC (surgery, radiation and chemotherapy) and no treatment.
Different SEER registry groupings are used to maximize the sample size. In the analysis of patients' characteristics and clinical-pathological features, SEER 9 data is applied and the following information is included: cancers diagnosed between 1973 and 2014, gender, marital status, age at diagnosis, age group (0-39 years, 40-64 years, 65-84 years, 85+ years), survival time, histological types, grade, stage (II-IV), treatment, tumor size, lymph node site (nodal, extra-nodal), and location. SEER 13 data containing detailed race and incidence information for cancers diagnosed between 1992 and 2014, is used in the analysis of incidence rate. Cancer diagnosed between 1973 and 2008 with follow-ups up to the end of 2013 from SEER 18 data is analyzed in the survival analysis.
Statistical analysis. This study uses Chi-square tests and ANOVA analysis in order to compare the patients' characteristics and clinical pathologic features across racial groups by SAS version 9.4. The age-standardized incidence rates and five-year relative survival rates are calculated by SEER* Stat software using US 2000 Census data. Multivariate Cox regressions are also performed when adjusting for gender, age, histological type, stage, location and treatment strategy. Incidence rates. The overall age-adjusted incidence rate is 19.53 per 100,000 person-years (Table 2).

Data availability.
Incidence rates increase with age. The four age groups have incidence rates 0.43, 15.36, 110.28, 154.23, respectively. Overall, non-Hispanic whites have the highest incidence rate, and the incidence rate is far greater than those of the other races. For the four age groups, non-Hispanic whites have the highest incidence rate, followed by blacks. APIs have the lowest rate. Overall, male have much higher incidence rate (34.46) than female (8.62). For both male and female, NHWs have the highest incidence rate, followed by blacks and then HWs. For the four major histological types, PTCC has the highest incidence rate, followed by NPTCC and then carcinoma. For PTCC and NPTCC, NHWs have the highest incidence rate (16.14 and 5.59, respectively), followed by blacks. While for Carcinoma, non-Hispanic whites and blacks have the same highest incidence rate (0.32). For SCC, blacks have the highest incidence rate (0.33), followed by NHWs (0.25). For all the four major histological types, APIs have the lowest incidence rate.
Survival rates. Figure 1 shows the unadjusted survival rate for five years. Non-Hispanic whites have better survival rates at all times, whereas the survival curves of other racial groups may cross. Blacks have a much lower survival rate than other racial groups. The five-year relative survival rates, stratified by gender, age group, histological types, stage, location, and treatment strategies are shown in Discussion. Bladder cancer is the most common malignancy of the urinary system. It is also one of the top ten common tumors of the body. Current studies investigated racial differences for a small number of racial groups, most of them focus on blacks and whites only 20,21 . Some other studies just cared about a small number of outcome variables 22 . In this study, UBC patients in the United States among non-Hispanic whites, Hispanic whites, blacks and APIs are compared comprehensively on the basis of SEER data. Significant racial differences are found in terms of patients' characteristics, clinical-pathologic features, incidence and survival. Such results can be informative for cancer epidemiologists, clinicians, and policy-makers. The observed gender, age, and tumor grade distributions are similar to those in the literature 18,21 . The statistical analyses suggest the existing of racial differences on the distributions of gender, marital status, age, age at diagnosis, survival time, histological types, grade, stage, treatment, tumor size, regional nodes negative, and tumor location. These findings are mostly consistent with those in the literature 23,24 . It is possible that the complex interactions of genetic makeup, living habits, occupational exposures, and environmental exposures cause the racial differences among UBC patients 25 .
Racial differences in incidence have been studied in a few publications 6,20 . Chen and others limited their study to blacks and whites, analyzed SEER data from 2010 to 2013, and found that blacks have lower incidence rates than whites 6 . Our analysis includes much wider racial distributions and generates more comprehensive results. Non-Hispanic whites have higher incidence rates of UBC than blacks while Hispanic whites have lower incidence rates than blacks and APIs have lower incidence rates than other races. Multiple factors may have contributions to the racial differences in incidence. One possible risk factor is smoking. As some studies reported that smoking significantly elevated bladder cancer risk (odds ratio = 2.4) and there are some known bladder carcinogens just like 2-napthylamine in tobaccos [26][27][28] . Non-Hispanic whites have been found that have a higher proportion of smokers compared with Hispanic whites, blacks, and Asian American 28 . The racial difference in UBC can also be attributed to occupational exposures. Numerous studies have demonstrated that workers in high-risk occupations involving exposure to dyes, rubber, leather, ink, or paints have an increased risk of bladder cancer. Three of the chemicals used in these industries (2-naphthylamine, benzidine, and 4-aminobiphenyl) are classified as human carcinogens by the International Agency for Research on Cancer (IARC) 29,30 . Some researchers already found the racial differences between non-Hispanic whites, Hispanic whites and blacks in exposure to occupational hazards 31 . Coffee drinking has been studied extensively as a potential risk factor, but the inconsistency of the observed association suggests that the relationship is quite weak 11 . In our study, we also found that when stratified by histological types, for SCC and Carcinoma, blacks have the highest incidence rate. Kantor and others also found that blacks have the highest incidence rate of SCC 32 .The high incidence of SCC appears to be related to a high incidence of endemic schistosomiasis in blacks 33 . Eradication of schistosomiasis can effectively reduce the incidence of SCC in blacks. It's not clear why blacks have the highest incidence of Carcinoma: further studies are needed.
The racial difference for survival outcome of UBC patients has been studied in some publications. Scosyrev and others compared the mortality rates of blacks and whites adjusted for grade, stage, and cell type in 2009, and found that for a time less than three years since diagnosis, blacks have higher mortality rates than whites 18 . Hollenbeck and others used SEER data for the years from 1992 through 2002 and found that blacks have an approximately 70% greater risk of cancer-related death compared with whites 21 . In our analysis, the survival advantage of whites is observed and blacks have the worst survival rate. The results are the same as those mentioned above. When stratified by treatment strategies, racial differences still exist for SR, SRC, and surgery, with blacks having the lowest survival rate. So we think that treatment may not be the main reason for the racial difference in survival rate. Hollenbeck's research already showed that the use of aggressive therapies (cystectomy, systemic chemotherapy, radiation) was found to be similar among black patients and white patients. After adjusting for the difference in treatment intensity and provider effects, the hazard ratios for blacks and whites just decrease 0.01 21 . Racial differences could also be observed from our study after stratified by stage and histological types, with blacks still have the worst survival rate. These findings support a recent SEER study that reported an excess   34 because people are more likely to have unhealthy lifestyles such as smoking, physical inactivity, and consumption of calorie-dense food in less developed areas 35 . Racial differences in survival may also be the end result of disparities in the healthcare provided to different races. Blacks may receive a lower quality of care relatively, which can reflect differential treatment by race.
In this study, we also found that APIs have the highest survival rate instead of whites in some cases. For histological types NPTCC and SCC, APIs have the highest survival rate. The possible reason may be different treatment strategies for some specific histological type in different racial groups. Additionally, different genetic makeup and lifestyle behaviors may contribute to this result. For treatment strategies SC, Chemotherapy, SRC, and SR, APIs have the highest survival rate than other racial groups. It reminds us that APIs may be more suitable for these treatment strategies mentioned above than other racial groups. This result can provide a reference for choosing the appropriate treatment strategies for different racial groups. The survival advantage of APIs is observed for stage II and overlapping lesion of bladder. Reasons for that need further researches. We can also learn from our study that tumor location in urachus and overlapping lesion of bladder has a lower survival rate than other locations. Tumor locations of UBC have rarely been studied before. It can provide a reference for prevention and treatment of bladder cancer.

Limitations.
The SEER database is analyzed in this study because it is the current largest cancer registry in the United States. However, it still has many limitations. Firstly, errors may arise in tumor classification and staging in multiple sites. But we do not expect a series of systematic errors correlated with ethnicity. Secondly, data collected in SEER may not be comprehensive enough. For instance, there are no data of genetic factors or lifestyle that may contribute to UBC. Furthermore, SEER does not collect data regarding insurance type, educational level, socioeconomic status, availability to healthcare and other factors that affect one's treatment, and as a consequence, the survival outcome. Additionally, the details about treatments are not recorded in SEER, such as surgical methods, chemotherapy dosage and so on. All above unmeasured patient differences may confound correlations between race and outcomes. Also, the study on SEER data is focus on patients from the United States only. It is not clear whether the results hold for other nations. In addition, although the sample size is at the same level as in related studies 18,36 , it would possible be over powered.

Conclusion
This study shows that there exist significant racial differences in patient characteristics, incidence and survivals among UBC patients in the U.S. The exact cause of such difference is not revealed. The identified racial difference could be informative and leveraged by UBC epidemiologists and clinicians despite the limitations.