Racial Disparities and Socioeconomic Status in Men Diagnosed With Testicular Germ Cell Tumors

Transcription

1 Racial Disparities and Socioeconomic Status in Men Diagnosed With Testicular Germ Cell Tumors A Survival Analysis Maxine Sun, BSc 1 ; Firas Abdollah, MD 1,2 ; Daniel Liberman, MD 1,3 ; Al a Abdo, MD 1,3 ; Rodolphe Thuret, MD 1,4 ; Zhe Tian, BSc 1 ; Shahrokh F. Shariat, MD 5 ; Francesco Montorsi, MD 2 ; Paul Perrotte, MD 4 ; and Pierre I. Karakiewicz, MD 1,4 BACKGROUND: Previous reports indicated that African-American men with testicular germ cell tumors (TGCTs) have more aggressive tumor characteristics and less favorable outcomes than other men. The authors of this report evaluated the effects of race and socioeconomic status (SES) on stage distribution, overall mortality (OM), and cancerspecific mortality (CSM) in men with TGCTs. METHODS: The Surveillance, Epidemiology, and End Results (SEER) database was used to identify 22,553 men who were diagnosed with TGCTs between 1988 and Kaplan-Meier and Cox regression analyses were generated to predict OM and CSM. Covariates of the analyses included race, SES, age, histologic subtype, disease stage, procedure type, SEER registry, and year of diagnosis. The interaction between race and SES also was examined. RESULTS: Overall, there were 516 African-American men, 21,090 Caucasian men, and 947 men of other races. African-Americans (14.9%) and individuals with low SES (10.7%) had a higher proportion of distant stage disease. CSM and OM rates were significantly higher for African-American patients and for patients who resided in low SES counties. Multivariate analyses revealed that African-American men and men with low SES were more likely to die of OM and CSM relative to Caucasian men (P <.001) and men with high SES (P <.001), respectively. The interaction between race and SES was not significant. CONCLUSIONS: African-American race and low SES appeared to predispose men to more advanced disease stages and to higher OM and CSM rates. These observations may warrant race-specific and/or SES-specific adjustments in the treatment of TGCT. Cancer 2011;117: VC 2011 American Cancer Society. KEYWORDS: race, socioeconomic status, testicular cancer, oncologic outcome, disparities. In multiple tumor sites, African-American race represents a risk factor for more advanced stage at diagnosis, more aggressive disease characteristics, and more unfavorable treatment outcomes Among patients with testicular cancer, few reports have examined the effect of race on disease stage and treatment outcomes Of the studies that did investigate race, 3 examined survival, which represents the ultimate treatment outcome for any tumor site All 3 studies demonstrated that African-American race was associated with worse survival in patients who were diagnosed with testicular cancer. However, all 3 studies relied on small populations of African-Americans (range, patients), and the most recent study focused on outcomes up to the Year Moreover, the effect of African-American race commonly is embedded within socioeconomic status (SES). However, none of the previous studies accounted for their plausible correlation. On the basis of these considerations, we decided to examine the contemporary effect of race on disease stage distribution and on mortality. The rationale for performing such analyses relied on potential improvements in the treatment and therapeutics of testicular cancer among African-American individuals over the past decade. Because race and SES are clearly related, we further extended our analysis to include SES and its interaction with race. Our analysis relied on a large, Corresponding author: Maxine Sun, BSc, Cancer Prognostics and Health Outcomes Unit, University of Montreal Health Center (CHUM), 1058, rue St-Denis, Montréal, Québec, Canada, H2X 3J4; Fax: (514) ; mcw.sun@umontreal.ca 1 Cancer Prognostics and Health Outcomes Unit, University of Montreal Health Center, Montreal, Quebec, Canada; 2 Department of Urology, Vita Salute San Raffaele University, Milan, Italy; 3 Department of Urology, University of Montreal Health Center, Montreal, Quebec, Canada; 4 Department of Urology, University of Montpellier Health Center, Montpellier, France; 5 Department of Urology, Weill Medical College of Cornell University, New York The first 2 authors contributed equally to this article. DOI: /cncr.25969, Received: October 6, 2010; Revised: November 18, 2010; Accepted: December 29, 2010, Published online March 8, 2011 in Wiley Online Library (wileyonlinelibrary.com) Cancer September 15,

2 contemporary ( ), population-based cohort of individuals with TGCT (n ¼ 22,553) that included 516 African-Americans. MATERIALS AND METHODS Study Population The study cohort originated from the Surveillance, Epidemiology, and End Results (SEER) Program, which is funded by the National Cancer Institute. Patient demographics, primary tumor site, tumor morphology and disease stage at diagnosis, first course of treatment, and follow-up of vital status are queried routinely from 17 SEER cancer registries. These consist of the San Francisco-Oakland, Detroit, Seattle Puget-Sound, Atlanta, San-Jose Monterey, Los Angeles, and Georgia metropolitan areas as well as the states of Connecticut, Hawaii, Iowa, New Mexico, Utah, Alaska, Greater California (excluding San Francisco, Los Angeles, and San Jose), Kentucky, Louisiana, and New Jersey. In general, the SEER population is considered to have a higher level of affluence, has lower unemployment rates, and is considered more urban than the rest of the United States population. 17 Up to the Year 2000, the SEER database covered approximately 10% of the US population, and it covered 26% thereafter. 18 We relied on the International Classification of Diseases ninth edition codes 9061 through 9063, 9070 through 9071, 9080 through 9085, and 9100 through 9102 to identify patients aged 18 years with TGCT in all stages with seminoma or nonseminoma histologic subtype. These criteria resulted in the identification of 22,826 patients. Further exclusions consisted of unknown stage (n ¼ 49), autopsy-diagnosed patients (n ¼ 24), and missing race (n ¼ 200). This resulted in a total of 22,553 assessable patients. Description of Covariates In the SEER database, race is self-reported by the patient at diagnosis and is defined as white, black, Asian (Chinese, Filipino, Japanese), American Indian/Alaska Native, and other unspecified. Race ascertainment in SEER is considered highly accurate and representative of the United States For our analyses, race was categorized as white, black, and other in the current study. Age was divided into tertiles (29 years vs years vs 40 years), and year of diagnosis was divided into quartiles ( vs vs vs ). The type of procedure was defined as orchiectomy alone versus orchiectomy with retroperitoneal pelvic lymph node dissection (RPLND) versus neither orchiectomy nor RPLND. Chemotherapy status was not available for analyses. Histologic subtype (seminoma vs nonseminoma), SEER stage (localized vs regional vs distant), and SEER registry of residence also were included. SES was determined using 3 ecologic attributes provided by the SEER Program: median family income, percentage of individuals living below the poverty line, and percentage of individuals without a high school diploma. Subsequently, a composite score was created using the sum of the standardized scores from the 3 variables, as described previously. 22,23 High and low SES groups were established by dividing the population according to the median value of the composite score. Statistical Analysis The chi-square test and the analyses of variance were used to compare proportions and means, respectively. Kaplan- Meier plots were constructed to graphically depict overall mortality (OM)-free and cancer-specific mortality (CSM)-free survival curves. Life tables were used to compute survival rates at specific time points. Subsequently, separate univariate and multivariate Cox regression analyses for predicting OM and CSM were fitted. Tested variables consisted of race categories, age groups, SES, histologic subtype, type of intervention, SEER stage, SEER registry, and year of diagnosis quartile. The interaction between SES and race was tested within multivariate analyses. Because seminoma and nonseminoma TGCTs behave differently, we repeated the above analyses according to histologic subtype. All statistical analyses were performed using the R statistical system (R Foundation for Statistical Computing, Vienna, Austria) and Predictive Analytics Software (PASW Statistics, version 17.0; SPSS Inc., Chicago, Ill). All P values were 2-sided, and statistical significance was set at P <.05. RESULTS Baseline characteristics are provided in Table 1. Overall, 22,553 patients were identified with TGCT of all stages. The majority of patients had seminomatous TGCT (60.4%). Most underwent orchiectomy without RPLND (83.4%). Of the patients with seminoma, 97.7% underwent orchiectomy alone versus 1.1% who underwent orchiectomy with RPLND. Of the patients with nonseminoma TGCT (39.6%), 87% versus 11.3% underwent 4278 Cancer September 15, 2011

3 Racial Disparities in Testicular Cancer/Sun et al Table 1. Descriptive Characteristics of 22,553 Men Who Were Diagnosed With Testicular Germ Cell Tumors Between the Years 1988 and 2006 No. of Men (%) Variable Overall, n522,553 (100) Caucasians, n521,090 (93.5) African-Americans, n5516 (2.3) Others, n5947 (4.2) P a Age group, y (32.5) 6819 (32.3) 156 (30.2) 360 (38) (37.5) 7899 (37.5) 209 (40.5) 350 (37) (30) 6372 (30.2) 151 (29.3) 237 (25) Histology Seminoma 13,620 (60.4) 12,697 (60.2) 335 (64.9) 588 (62.1).05 Nonseminoma 8933 (39.6) 8393 (39.8) 181 (35.1) 359 (37.9) Type of procedure Orchiectomy 21,075 (83.4) 19,701 (93.4) 486 (94.2) 888 (93.8).002 Orchiectomy and RPLND 1155 (5.1) 1096 (5.2) 14 (2.7) 45 (4.8) None 323 (1.4) 293 (1.4) 16 (3.1) 14 (1.5) Tumor SEER stage Localized 16,251 (72.1) 15,251 (72.3) 334 (64.7) 666 (70.3) <.001 Regional 4113 (18.2) 3839 (18.2) 105 (20.3) 169 (17.8) Distant 2189 (9.7) 2000 (9.5) 77 (14.9) 112 (11.8) Socioeconomic status High 11,201 (49.7) 10,474 (49.7) 156 (30.2) 571 (60.3) <.001 Low 11,352 (50.3) 10,616 (50.3) 360 (69.8) 376 (39.7) SEER registries California 3205 (14.7) 3167 (15) 39 (7.6) 99 (10.5) <.001 Los Angeles 3024 (13.5) 2827 (13.4) 76 (14.7) 131 (13.8) Seattle 2322 (10.3) 2249 (10.7) 18 (3.5) 55 (5.8) San Francisco 2000 (8.9) 1797 (8.5) 55 (10.7) 148 (15.6) Detroit 1772 (7.9) 1685 (8) 79 (15.3) 8 (0.8) Connecticut 1633 (7.2) 1583 (7.5) 28 (5.4) 22 (2.3) New Jersey 1508 (6.7) 1410 (6.7) 59 (11.4) 39 (4.1) Iowa 1409 (6.2) 1395 (6.6) 7 (1.4) 7 (0.7) Utah 1121 (5) 1100 (5.2) 1 (0.2) 20 (2.1) Atlanta 993 (4.4) 909 (4.3) 71 (13.8) 13 (1.4) San Jose 928 (4.1) 855 (4.1) 8 (1.6) 65 (6.9) New Mexico 777 (3.4) 719 (3.4) 0 (0) 58 (6.1) Kentucky 630 (2.8) 615 (2.9) 15 (2.9) 0 (0) Louisiana 569 (2.5) 513 (2.4) 52 (10.1) 4 (0.4) Hawaii 499 (2.2) 254 (1.2) 4 (0.8) 241 (25.4) Alaska 37 (0.2) 0 (0) 0 (0) 37 (3.9) Rural Georgia 16 (0.1) 12 (0.1) 4 (0.8) 0 (0) Year of diagnosis, quartiles (25.8) 5470 (25.9) 106 (20.5) 252 (26.6) (24.7) 5180 (24.6) 134 (26) 250 (26.4) (24.6) 5193 (24.6) 155 (30) 200 (21.1) (24.9) 5247 (24.9) 121 (23.4) 245 (25.9) RPLND indicates retroperitoneal pelvic lymph node dissection; SEER, Surveillance, Epidemiology, and End Results Program of the National Cancer Institute. a P values were determined with analyses of variance and chi-square tests. orchiectomy alone versus orchiectomy with RPLND, respectively. A small proportion of patients (1.4%) did not undergo any surgical intervention. Most patients originated from the Greater California (14.7%) and Los Angeles (13.5%) SEER registries, whereas the fewest originated from the Rural Georgia SEER registry (0.1%). The median age was 32 years. The length of follow-up for the patients who did not die during the study period was 8.4 years (median, 7.0 years). After stratification according to race group, there were 516 African-American men (2.3%), 21,090 Caucasian men (93.5%), and 947 men of other races (4.2%) Cancer September 15,

4 (Table 1). The average age was 35 years (median, 34 years), 35 years (median, 34 years), and 34 years (median, 33 years),respectively,forthesameracegroups(p <.001). Stratification of disease stage according to race for African-Americans, Caucasians, and other races revealed that the proportion of individuals with regional disease (20.3% vs 18.2% vs 17.8%, respectively) and distant disease (14.9% vs 9.5% vs 11.8%, respectively) was highest among African-Americans (P <.001) relative to Caucasians and individuals of other races. Stratification of disease stage according to low SES versus high SES revealed that the proportion of individuals with regional disease (18.6% vs 17.9%, respectively) or distant disease (10.7% vs 8.7%, respectively) was highest among patients residing in low SES counties (P <.001) relative to patients residing in high SES counties. Stratification of procedure type according to race in patients with nonseminoma TGCT revealed that the rate of orchiectomy with RPLND was lowest in African- Americans compared with Caucasians and individuals of other races (6.6% vs 11.5% vs 9.5%, respectively; P ¼.2). Stratification of procedure type according to SES in the same subgroup revealed that the rate of RPLND after orchiectomy was lower in patients residing in low SES counties relative to patients residing in high SES counties (9.3% vs 13.4%, respectively; P <.001). Stratification of histologic subtype according to race for African-Americans, Caucasians, and other races revealed that the rate of nonseminoma TGCT was similar across race categories (35.1% vs 39.8% vs 37.9%, respectively; P ¼.05). Conversely, after stratification of histologic subtype according to SES, the rate of nonseminoma TGCT was higher in patients residing in low SES counties compared with patients residing in high SES counties (40.8% vs 38.4%, P <.001). After stratification of SES according to race for African-Americans, Caucasians, and other races, more African-Americans resided in low SES counties than Caucasians and other races (69.8% vs 50.3% vs 39.7%, respectively; P <.001). It is interesting to note that most African-Americans originated from the Detroit SEER registry (15.3%), whereas the majority of Caucasians and individuals of other races originated from the Greater California (15%) and Hawaii (25.4%) SEER registries, respectively (P <.001). Finally, the rates of African- Americans ( : 1.9%-2.2%; P ¼.1), Caucasians ( : 93%-93.5%; P ¼.6), and patients of other races ( : 4.3%-4.4%; P ¼.5) remained stable over time. Figure 1. These charts illustrate the rates of (A) overall mortality (OM)-free survival and (B) and cancer-specific mortality (CSM)-free survival for the entire population. CI indicates confidence interval. Kaplan-Meier plots and life tables revealed that 5.1% (95% confidence interval [CI], 4.8%-5.4%) of all patients with testicular tumors died of any cause 5 years after the start of follow-up (Fig. 1A). The overall 5-year CSM rate was 2.8% (95% CI, 2.6%-3.1%) (Fig. 1B). When mortality was stratified according to race, the 5- year OM rate for African-Americans was 10.7% (95% CI, 7.8%-13.6%) versus 4.9% for Caucasians (95% CI, 4.5%-5.2%) and 7.6% for patients of other races (95% CI, 5.8%-9.4%; log-rank P <.001) (Fig. 2A). The 5-year 4280 Cancer September 15, 2011

5 Racial Disparities in Testicular Cancer/Sun et al Figure 2. These charts illustrate the rates of (A) overall mortality (OM)-free survival and (B) and cancer-specific mortality (CSM)-free survival stratified according to race groups. CI indicates confidence interval. CSM rate for African-Americans was 6.7% (95% CI, 4.3%-9.0%) versus 2.7% for Caucasians (95% CI, 2.5%-2.9%) and 3.5% for patients of other races (95% CI, 2.2%-4.8%; log-rank P <.001) (Fig. 2B). When mortality was stratified according to SES, the 5-year OM rate for low SES was 6.1% (95% CI, 5.6%-6.5%) versus 4.2% (95% CI, 3.8%-4.6%) for high SES (log-rank P <.001) (Fig. 3A), and the 5-year CSM rates were 3.4% (95% CI, 3.1%-3.8%) versus 2.2% (95% CI, 1.9%-2.5%) for the same groups, respectively (log-rank P <.001) (Fig. 3B). After adjusting for all covariates, African-Americans were 1.6-fold more likely to die of OM than their Caucasian counterparts (P ¼.001) (Table 2). Similarly, the Figure 3. These charts illustrate the rates of (A) overall mortality (OM)-free survival and (B) and cancer-specific mortality (CSM)-free survival stratified according to socioeconomic status groups. CI indicates confidence interval. effect of race on CSM indicated that African-Americans had a 2.1-fold greater risk of CSM in multivariate analyses (P <.001). Low SES portended a 1.3-fold greater risk of OM than high SES (P <.001). Compared with men residing in high SES counties, men residing in low SES counties were 1.4-fold more likely to die of CSM (P <.001). The test for interaction between race and SES failed to achieve independent predictive status in OM and CSM analyses (data not shown). In patients with seminoma TGCT, African-American race was not associated significantly with a higher rate of OM (hazard ratio [HR], 1.1; P ¼.8) or CSM (HR, 1.3; P ¼.5). Men with low SES had a higher rate of OM (HR, 1.4; P ¼.003) and CSM (HR, 1.7; P ¼.007). By comparison, African-Americans with nonseminoma Cancer September 15,

6 Table 2. Univariate and Multivariate Cox Regression Analyses for Predicting Overall Mortality and Cancer-Specific Mortality in the Entire Population Overall Mortality Cancer-Specific Mortality UVA MVA UVA MVA Predictor HR P HR P HR P HR P Race category Caucasian (Ref) African-American 2.20 < < <.001 Other Socioeconomic status High (Ref) Low 1.36 < < < Tumor SEER stage Localized (Ref) Regional 1.96 < < < <.001 Distant 8.53 < < < <.001 Histology Seminoma (Ref) Nonseminoma 1.85 < < < <.001 Type of procedure Orchiectomy (Ref) Orchiectomy and RPLND < <.001 No intervention 7.92 < < < <.001 Age group, y 29 (Ref) < < <.001 SEER registries California (Ref) Los Angeles Seattle San Francisco Detroit Connecticut New Jersey Iowa Utah Atlanta San Jose New Mexico Kentucky Louisiana Hawaii Alaska Rural Georgia NA NA NA NA Year of diagnosis, quartile (Ref) < < UVA indicates univariate analysis, MVA: multivariate analysis, HR, hazard ratio; Ref, referent category; SEER, Surveillance, Epidemiology and End Results Program of the National Cancer Institute; RPLND, retroperitoneal pelvic lymph node dissection; NA, not available because of insufficient number of observations. TGCT had higher rates of OM (HR, 2.4; P <.001) and CSM (HR, 2.5; P <.001). Also, patients with nonseminoma TGCT who resided in low SES counties had a higher rate of CSM (HR, 1.3; P ¼.04) but not OM (HR, 1.2; P ¼.08). All examined interactions in this subanalysis failed to achieve statistical significance Cancer September 15, 2011

7 Racial Disparities in Testicular Cancer/Sun et al DISCUSSION Testicular cancer accounts for approximately 1% of all malignancies among men and primarily affects men between ages 15 years and 34 years. 24 The incidence of testicular cancer remains the highest among Caucasians within the United States. Conversely, the incidence of testicular cancer in African-Americans is very low. However, several authors previously reported an important increase in the incidence of testicular cancer among African-Americans. 21,25,26 A recent examination using the SEER database of race-specific incidence suggested that rates of TGCT among African-American men increased by 70.4% from 1973 to This increase represented the highest annual percentage change (þ2.3%; P <.05) relative to Caucasians and other races. 26 Authors have surmised that, although reasons for the rising rates are unknown, because of a lack of increase in the proportion of tumors diagnosed at localized stage in African-Americans, screening and earlier diagnosis of TGCT do not seem to explain the phenomenon. 21 Although the incidence of TGCT in African-Americans is rising, data examining mortality rates among African-Americans are scarce. Despite the relevance of these comparative studies, their sample sizes (n ¼ 53, n ¼ 326, and n ¼ 329) limit generalizability of the findings Moreover, the most recent of those reports examined mortality trends only up to the Year Taken together, it appears that the effect of race deserves a reappraisal within a larger and more contemporary patient cohort. In addition, based on the close association between race and SES, 8,9,27 the effect of SES on disease stage and treatment outcomes 16,27,28 and the interaction between race and SES deserve a formal analysis. Our study focused on these objectives. To obviate the limitations of previous analyses, we decided to rely on a large, contemporary, populationbased cohort from the United States. The query was performed on the National Cancer Institute-sponsored SEER Program database. In total, 22,553 patients who were diagnosed with TGCT between 1988 and 2006 were identified. African-Americans represented 2.3% of this population (n ¼ 516). The results indicated that African-American men had a higher rate of regional and metastatic disease than their Caucasian counterparts. A similar effect was recorded for SES, in which low SES was associated with a higher rate of regional and distant stages. In the subset of patients with nonseminoma TGCT, the rate of RPLND did not vary significantly according to race; conversely, it was lower in patients who resided in low SES counties. In addition, OM and CSM rates were highest in African-American men, even after adjusting for all covariates. Analyses of the interaction term between race and SES failed to reach a statistical effect, thereby confirming that racial disparities were not limited only to a specific SES category. It is noteworthy that, after stratifying patients according to histologic subtype, the effect of race disappeared in men who had seminoma TGCT. Conversely, the detrimental effect of African-American race on OM and CSM remained in men who had nonseminoma TGCT. Finally, patients who resided in low SES counties had a worse survival in all stratifications, and the only nonsignificant difference was observed for OM among men who had nonseminoma TGCT. Taken together, our findings indicate that race and SES predispose men to a more unfavorable disease stage and more unfavorable survival rates. Our findings are consistent with 2 previous reports on race-specific stage distribution. 12,14 Biggs and Schwartz 12 (n ¼ 329) observed that African-Americans were diagnosed at a later stage and demonstrated worse survival (HR, 2.3; P <.001). Similarly, Gajendran et al 14 (n ¼ 326) observed that African- Americans had a higher rate of distant stage disease (16% vs 12%) and also demonstrated worse survival (HR, 2.3; P <.001); however, after adjusting for histologic subtype and disease stage at presentation, differences in survival disappeared (HR, 1.4; P ¼.2). In a smaller study by Bridges et al (n ¼ 53), the authors did not observe any significant differences between clinical stage and race. 13 However, a worse survival for African-Americans also was noted in that study. In our analysis, even after adjusting for disease stage, African-Americans had higher mortality rates. Moreover, African-American race was an indicator of worse survival in analyses that were restricted to men with nonseminoma TGCT. Because the latter represents the more clinically aggressive tumor, this finding is worrisome: It suggests the possibility that tumor biology may differ between African-Americans and Caucasian. 8 Specifically, it has been postulated that African-American patients harbor a more aggressive phenotype of TGCT than their Caucasian counterparts. However, other investigators could not confirm this possibility in a meta-analysis. 29 This hypothesis warrants further testing in a database that contains more detailed tumor characteristics. Several explanations may be proposed to account for the observed race-specific and SES-specific differences observed in the current study and in previous reports. For example, race-specific perceptions of disease and a lack of understanding of cancer screening, notably the Cancer September 15,

8 importance of regular self-examinations, may contribute to delayed presentation. 14,21,25,30 Indeed, race-specific variations in cancer detection and staging evaluations may seem to increase survival rates among Caucasians versus African-Americans to a large extent through lead-time bias 31 and/or stage migration. 32 Although this may be a possible explanation for the increased survival probabilities reported among Caucasians relative to African-Americans, several other factors also may be involved. Similarly, several well designed epidemiologic studies tested the effects of SES and health insurance protection as risk factors The finding that low SES influences treatment outcomes has been investigated previously in other malignancies. Environmental, cultural, nutritional, and lifestyle factors as well as difficulty in accessing the healthcare system also may represent plausible explanations. Nonetheless, none of these variables fully accounts for the observed differences in survival according to race and/or SES. The current study raises important implications for future perspectives, including whether the observed survival differences were caused by disparities in treatment patterns or treatment screening. In addition, whether these disparities originate from differences in receipt of treatment, decision-making or follow-up, an improved health education and increased participation in screening may be indicated. Further studies will be needed to identify factors related to the increased mortality rates in African-American men with TGCT and individuals residing in low SES counties. Moreover, future studies also should examine whether survival differences exist in other communities, such as Hispanic and Asian communities. Our current study was not devoid of limitations, 1 of which was a lack of information on chemotherapy status and receipt of radiotherapy. A previous randomized phase 3 trial demonstrated superior recurrence-free survival among patients who received 1 course of bleomycin, etoposide, and cisplatin chemotherapy over RPLND. 36 Moreover, 7% of patients require adjuvant chemotherapy after RPLND. 37 Consequently, information on chemotherapy status would have been an important addition. It also is possible that patients who did not undergo RPLND may have opted for radiation therapy. However, information on chemotherapy and radiotherapy status was unavailable equally for all race and SES categories. Consequently, its availability or lack thereof had no differential effect across the examined strata and no bearing on the observed CSM and OM rates. Nevertheless, the absence of information on chemotherapy and radiotherapy status meant that we were unable to comment on specific cancer control outcomes related to the use of such treatment modalities. In addition, we were unable to account for the time interval between diagnosis and treatment. The disparities in race and SES can be embedded within disparities in healthcare access, physician-specific differences, and/or patient noncompliance with follow-up. Information on time to treatment would have provided valuable insight into plausible explanations for the observed disparities and possibly raised implications for healthcare equity. Moreover, the SES measure represents an ecologic combined variable, which reflects the SES of the county of residence of an individual and not of the individual himself. An overestimate or underestimate of the effect of SES on survival may be plausible. Although this method has been established in previous studies, 22 it may not be the most accurate SES measure. Consequently, the detrimental effect of SES on survival observed in the current study should be interpreted with caution. Future studies that examine the effect of SES on survival should attempt to measure SES on an individual level. Other limitations include the lack of information on methods of cancer detection and health insurance status, because the improved survival among Caucasians and individuals with high SES may have been caused by lead-time bias and/or stage migration. Moreover, lack of information on comorbidity status and recurrence data limited the interpretation of our findings. These limitations also were apparent in previous reports that addressed the same topic In conclusion, African-Americans in the current study had a higher rate of metastatic disease. Higher mortality rates in African-Americans relative to Caucasians, even after adjusting for disease stage and histologic subtype, suggest that delayed disease stage at presentation may not completely explain the difference. Further studies are needed to assess the underlying factors. CONFLICT OF INTEREST DISCLOSURES Dr. Pierre I. Karakiewicz is partially supported by the University of Montreal Urology Specialists, Fonds de la Recherche en Santé du Québec, the University of Montreal Department of Surgery, and the University of Montreal Foundation. REFERENCES 1. Chavez-Macgregor M, Litton J, Chen H, et al. Pathologic complete response in breast cancer patients receiving anthracycline- and taxane-based neoadjuvant chemotherapy: 4284 Cancer September 15, 2011

9 Racial Disparities in Testicular Cancer/Sun et al evaluating the effect of race/ethnicity. Cancer. 2010;116: Du XL, Liu C-C. Racial/ethnic disparities in socioeconomic status, diagnosis, treatment and survival among Medicareinsured men and women with head and neck cancer. J Health Care Poor Underserved. 2010;21: Fedewa SA, Ward EM, Stewart AK, et al. Delays in adjuvant chemotherapy treatment among patients with breast cancer are more likely in African-American and Hispanic populations: a national cohort study J Clin Oncol. 2010;28: Hollenbeck BK, Dunn RL, Ye Z, et al. Racial differences in treatment and outcomes among patients with early stage bladder cancer. Cancer. 2010;116: Scosyrev E, Noyes K, Feng C, et al. Sex and racial differences in bladder cancer presentation and mortality in the United States. Cancer. 2009;115: Vicini F, Jones P, Rivers A, et al. Differences in disease presentation, management techniques, treatment outcome, and toxicities in African-American women with early stage breast cancer treated with breast-conserving therapy. Cancer. 2010;116: Shavers V, Brown M. Racial and ethnic disparities in the receipt of cancer treatment. J Natl Cancer Inst. 2002;94: Albain KS, Unger JM, Crowley JJ, et al. Racial disparities in cancer survival among randomized clinical trials patients of the Southwest Oncology Group. J Natl Cancer Inst. 2009;101: Artinyan A, Mailey B, Sanchez-Luege N, et al. Race, ethnicity, and socioeconomic status influence the survival of patients with hepatocellular carcinoma in the United States. Cancer. 2010;116: Pettaway C. Racial differences in the androgen/androgen receptor pathway in prostate cancer. J Natl Cancer Inst. 1999;91: Bach P, Cramer L, Warrent J, et al. Racial differences in the treatment of early-stage lung cancer. N Engl J Med. 1999;341: Biggs ML, Schwartz SM. Differences in testis cancer survival by race and ethnicity: a population-based study, (United States). Cancer Causes Control. 2004;15: Bridges P, Sharifi R, Razzaq A, et al. Decreased survival of black Americans with testicular cancer. J Urol. 1998; 159: Gajendran V, Nguyen M, Ellison L. Testicular cancer patterns in African-American men. Urology. 2005;66: Moul J, Schanne F, Thompson I, et al. Testicular cancer in blacks. Cancer. 1993;73: Hoffman KE, Chen M-H, Punglia RS, et al. Influence of year of diagnosis, patient age, and sociodemographic status on recommending adjuvant radiation treatment for stage I testicular seminoma. J Clin Oncol. 2008;26: Nattinger AB, McAuliffe TL, Schapira MM. Generalizability of the Surveillance, Epidemiology, and End Results registry population: factors relevant to epidemiologic and health care research. J Clin Epidemiol. 1997;50: National Cancer Institute; Surveillance, Epidemiology, and End Results (SEER) Program. About the SEER Program. Bethesda, MD: National Cancer Institute; Accessed September Bach P, Guadagnoli E, Schrag D, et al. Patient demographic and socioeconomic characteristics in the SEER-Medicare database: applications and limitations. Med Care. 2002; 40:IV19-IV US Census Bureau. Census Washington, DC: US Census Bureau; McGlynn KA, Devesa SS, Graubard BI, et al. Increasing incidence of testicular germ cell tumors among black men in the United States. J Clin Oncol. 2005;23: Du XL, Fang S, Coker AL, et al. Racial disparity and socioeconomic status in association with survival in older men with local/regional stage prostate carcinoma: findings from a large community-based cohort. Cancer. 2006;106: Robert SA, Strombom I, Trentham-Dietz A, et al. Socioeconomic risk factors for breast cancer: distinguishing individual- and community-level effects. Epidemiology. 2004;15: Jemal A, Siegel R, Ward E, et al. Cancer statistics, CA Cancer J Clin. 2009;59: McGlynn KA, Devesa S, Sigurdon A, et al. Trends in the incidence of testicular germ cell tumors in the United States. Cancer. 2003;97: Holmes L, Escalante C, Garrison O, et al. Testicular cancer incidence trends in the USA ( ): plateau or shifting racial paradigm? Public Health. 2008;122: Kaufman J, Cooper R, McGee D. Socioeconomic status and health in blacks and whites: the problem of residual confounding and the resiliency of race. Epidemiology. 1997; 8: Mackenbach J, Stirbu I, Roskam A, et al. Socioeconomic inequalities in health in 22 European countries. N Engl J Med. 2008;358: Bach P, Schrag D, Brawley O, et al. Survival of blacks and whites after a cancer diagnosis. JAMA. 2002;287: Klein JB, Nguyen CT, Saffore L, et al. Racial disparities in urologic health care. J Natl Med Assoc. 2010;102: Shwartz M. Estimates of lead time and length bias in a breast cancer screening program. Cancer. 1980;46: Feinstein A, Sosin D, Wells C. The Will Rogers Phenomenon: stage migration and new diagnostic techniques as a source of misleading statistics for survival in cancer. N Engl J Med. 1985;312: Marcella S, Miller J. Racial differences in colorectal cancer mortality: the importance of stage and socioeconomic status. J Clin Epidemiol. 2001;54: Ragland K, Selvin S, Merrill D. Black-white differences in stage-specific cancer survival: analysis of 7 selected sites. Am J Epidemiol. 1991;133: Steinhorn S, Myers M, Hankey B, et al. Factors associated with survival differences between black women and white women with cancer of the uterine corpus. Am J Epidemiol. 1986;124: Albers P, Siener R, Krege S, et al. Randomized phase III trial comparing retroperitoneal lymph node dissection with 1 course of bleomycin and etoposide plus cisplatin chemotherapy in the adjuvant treatment of clinical stage I nonseminomatous testicular germ cell tumors: AUO trial AH 01/ 94 by the German Testicular Cancer Study Group. J Clin Oncol. 2008;26: Nguyen CT, Fu AZ, Gilligan TD, et al. Defining the optimal treatment for clinical stage I nonseminomatous germ cell testicular cancer using decision analysis. J Clin Oncol. 2010;28: Cancer September 15,