In 1983, the National Cancer Institute began a socioepidemiologic study of possible biologic and behavioral influences on racial disparities in cancer presentation and mortality. Bladder cancer was 1 of the organ systems highlighted in the study. Twenty-five years later, the racial and sex differences in bladder cancer persist, and multiple population-based studies have begun to shed some light on this issue. In this issue of Cancer, Scosyrev et al from the University of Rochester have published a very informative and clinically relevant population-based analysis of sex and racial differences in bladder cancer.1 The authors used the Surveillance, Epidemiology, and End Results (SEER) database to confirm the findings of others that African-American (AA) patients and women present with more advanced stages of disease and have more aggressive histologies at the time of diagnosis. Moreover, as described previously, their current study reiterated the findings that men with bladder cancer have a better cancer-specific survival than women and that white patients have a better survival compared with AA patients. This is despite the finding that whites have a nearly 2 times higher incidence of bladder cancer than AA patients and men have a greater than 3-fold incidence compared with women.2
The novel and interesting findings of the study by Scosyrev et al revealed that, when controlling for prognostic factors such as age, stage, grade, and histologic subtype, the cancer-specific mortality still was increased significantly in women and AA patients. Although previous studies have demonstrated worse outcomes for women and AA patients with regard to bladder cancer, to our knowledge this is 1 of the first studies that controlled for multiple tumor characteristics and still quantified a significant sex and racial disparity. Although much of the findings could be explained by the unequal distribution of adverse tumor characteristics, this did not fully account for the differences in cancer-specific survival, because the hazards ratios remained significant after controlling for the covariates. These findings strongly suggest that other factors play a role in the inferior outcomes for these patient populations. The authors speculate that possible contributing factors include access to healthcare, delay in diagnosis and therapy, choice of treatment options, host factors, and differences within a tumor characteristic that were not detected within the SEER database (eg, tumor substaging).
The differences in cancer-specific mortality between sex and race were not consistent over time. The excess hazard of death from bladder cancer ranged from 80% to 100% for the first year of follow-up, decreased but remained elevated during the second year, and disappeared after the third or fourth years of follow-up. The authors point out that, although the differences in survival disappear after 3 to 4 years, most deaths from bladder cancer occur within the first 2 to 3 years. The likely reason for the convergence of the hazards ratios over time (ie, loss of significant differences in cancer-specific survival among the groups) is that most of the deaths occurred earlier in the follow-up period. Moreover, >50% of the patients in each group had stage I disease, which also decreased the overall number of cancer-related deaths in the study. Thus, the disparity in death from bladder cancer observed within the first 2 years of follow-up should have a significant impact on the overall differences in survival attributable to sex and race.
The sex and racial differences identified in the study by Scosyrev et al. are compelling and corroborate several previous population-based reports. Lee et al examined the trends in bladder cancer presentation and survival in white and AA patients over a 27-year period.3 They observed that AA patients consistently presented with higher stage and grade tumors over the entire study period. Disease-specific survival, controlling for stage and grade, also was significantly worse for AA patients with localized and locally advanced disease. On multivariate analysis, AA race was an independent predictor of inferior cancer-specific survival. The most impressive finding of that SEER-based study was the persistence of the racial disparity over 27 years. Another population-based study by Prout et al demonstrated that poorer overall survival for AA patients was limited to those with pathologic T2 (pT2) and pT3 disease.4 Similarly, in that study, AA patients presented with higher grade and stage tumors. However, AA patients and white patients did not differ with respect to diagnostic tests performed or therapy received. This is in contrast to earlier reports suggesting that survival differences between races were partly because AA patients were less likely to undergo surgery than white patients.5 The authors speculated that differences in comorbidities (not well differentiated in this study) and extent of disease within tumor stage may have had an impact on the differences in outcomes. Underwood et al examined the interaction between race, sex, and geographic location on the racial disparity in bladder cancer.6 Using the SEER database, those authors observed that not only did AA patients have more aggressive tumors and demonstrate poorer cancer-specific survival but also that women of both races had worse outcomes, and AA women had the worst survival. The differences were greatest in certain regions of the country, particularly the metropolitan Atlanta area.
Previous studies also have documented the sex disparity in bladder cancer survival. By using the SEER database, Mungan et al reported a stage-adjusted worse survival for women in the US.7 It is noteworthy that the second arm of that study examined the same survival outcomes in the Netherlands and identified less striking sex differences. Those findings, however, support the notion that differences in stage at presentation are unlikely to fully explain the worse prognosis for women with bladder cancer.
The etiologies of the sex and racial disparities in bladder cancer incidence and survival are yet to be elucidated. The current study by the Scosyrev group is timely, because it reiterates that clinicians must take the next steps in investigating racial and sex inequalities in bladder cancer. There is no question that the disparities exist, even when controlling for multiple tumor characteristics (eg, stage and grade). For the most part, however, researchers have been able only to speculate regarding why these differences exist. Recently, some studies have begun to shed more light on the etiology of the sex and racial differences in bladder cancer.
Taub et al examined the Nationwide Inpatient Sample from 1988 to 2000.8 Over 15,000 patients who underwent radical cystectomy in the US were included in their analysis. Both inpatient mortality and hospital length of stay (LOS) were increased significantly in AA patients compared with white patients, even when controlling for comorbidities, hospital factors, and demographics. Women also experienced a longer LOS compared with men. Inpatient mortality and LOS may be surrogate markers for the quality of surgical and postoperative care. Thus, the findings described above suggest unequal or inferior treatment as a possible cause of the sex and racial inequalities. Suboptimal treatment may be because of patient, physician, and/or system factors. Certain patients may deny the best treatment options available, especially when urinary diversion requiring a stoma and/or sexual dysfunction occurs. Provider diagnostic and therapeutic choices may be influenced by race and sex,9 and physicians may not provide the best treatment because of lack of knowledge or training and the extensive work required to care for patients with bladder cancer. Finally, system-level problems include minorities being enrolled in inferior health plans and having poorer access to healthcare.
A striking example of a provider-based etiology for sex disparity was demonstrated in a recent study by Johnson et al.10 They examined the referral patterns of hematuria to urologists and observed that men were significantly more likely than women to be referred for new onset or recurrent hematuria. Men had a 65% increased likelihood for urologic referral compared with women. Those authors speculated that a presumptive diagnosis of urinary tract infection in women could be the culprit. Nonetheless, a delay in urologic referral for hematuria certainly could result in a delay in diagnosis of life-threatening illnesses such as urothelial carcinoma. It is our impression that, all too often, bladder cancer is not even in the differential diagnosis when women present to primary care physicians complaining of hematuria or a change in voiding symptoms. Frequently, women are treated with antibiotics without urine analysis, urine cultures, or cytology.
Whether or not AA and women patients harbor biologically more aggressive cancers than other patient populations remains an important unanswered question. However, until research at the molecular level can help sort out genetic differences in bladder cancer, clinicians must focus on other social factors that influence the racial and sex disparity in this disease. The current study by Scosyrev et al elegantly demonstrates that, even after controlling for tumor characteristics, inferior outcomes remain for AA patients and women.1 Unfortunately, some of the major limitations of population-based studies are the lack of information on comorbid disease, socioeconomic status, access to healthcare, and delay in diagnosis. Future work should be performed on adding additional data fields to the SEER database to help answer these questions. Another major limitation of the large-scale, population-based studies described above is the lack of more precise staging. Although stage and grade are controlled for in the analyses, there still may be significant differences in extent of disease between 2 groups that are considered equal. Indeed, the stage breakdown is not even always by true pathologic or clinical T stage but, rather, by more general criteria, such as ‘localized,’ ‘regional,’ and ‘distant.’ Again, improving the detail of these population-based databases will help identify true differences between patient cohorts. The studies described above on referral patterns for hematuria and surgical outcomes and LOS for bladder cancer have begun to demonstrate a trend toward inferior quality of care for AA and women. However, no standardized framework to assess the quality of care for bladder cancer currently exists. Without a measurement tool, it will be more difficult to narrow the gap in racial and sex disparity. A quality-of-care infrastructure has been put into place for early-stage prostate cancer and could serve as a starting point for developing a similar system for bladder cancer.11 Greater public awareness and better medical student and primary care physician education also are critically important to improve the earlier diagnosis and to standardize treatment for all patients with bladder cancer. It is our impression that the majority of patients with bladder cancer have a delay in diagnosis because of a lack of urgency for urologic evaluation for hematuria. Earlier diagnosis is crucial for bladder cancer survival. Recently, groups such as the Bladder Cancer Advocacy Network (www.BCAN.org accessed October 22, 2008) and the American Bladder Cancer Society (www.bladdercancersupport.org accessed October 22, 2008) have been created to promote bladder cancer awareness.