We express our appreciation to the dedicated staff of the NCI Surveillance, Epidemiology, and End Results program and cancer registries contributing data to this program. We also thank Mr. David Check of the Division of Cancer Epidemiology and Genetics for figure development.
Patients with renal cell carcinoma (RCC) who are black tend to have poorer prognosis than similar patients who are white. This study examined whether the racial disparity in RCC patient survival varies by demographic and clinical characteristics.
Nearly 40,000 patients (4359 black and 34,991 white) diagnosed with invasive RCC from 1992 to 2007 were identified from 12 registries in the National Cancer Institute Surveillance, Epidemiology, and End Results program, covering approximately 14% of the US population. Relative survival rates through 2008 were computed using the actuarial method.
Proportionally more blacks than whites were diagnosed with RCC under age 50 and with localized cancer. Overall, the 5-year relative survival rates were 72.6% (95% confidence interval 72.0%-73.2%) for white and 68.0% (66.2%-69.8%) for black patients. Survival was higher among women than men and among younger than older patients. Survival decreased with advancing tumor stage and, within each stage, decreased with increasing tumor size. Patients with clear cell RCC, a more common form among whites, had poorer prognosis than patients with papillary or chromophobe subtypes, which are more common among blacks. Survival for patients who received no surgical treatment (10.5% of white patients and 14.5% of black patients) was substantially lower than for patients treated with nephrectomy, with similar survival among whites and blacks. In all other demographic and clinical subgroups of patients, whites consistently had a survival advantage over blacks.
Malignant tumors of the kidney and renal pelvis together rank eighth in cancer incidence and 12th in cancer mortality in the United States, with more than 64,770 new cases and 13,570 deaths estimated for 2012.1 More than 90% of these tumors are adenocarcinomas, originating in the renal parenchyma (renal cell carcinoma [RCC]).2 RCC incidence has been rising substantially for at least 4 decades, with more rapid increases and emerging higher rates among black than white Americans since the mid-1990s.3 Whereas rates for tumors of all stages and sizes have risen, the majority of the increases since the 1980s have occurred in localized tumors, especially those < 2.0 cm in size.2, 4 In contrast, mortality from kidney cancer has leveled since the 1990s, with similar death rates between black and white Americans.5
Tumor stage and size are well documented to influence cancer prognosis.6, 7 Tumor histological subtype has also been evaluated as a prognostic indicator for patients with RCC.8-11 Although papillary or chromophobe RCC have been shown to confer a better prognosis than clear cell RCC, the predominant subtype,9, 10, 12 histological subtype was not found to be an independent prognostic indicator in other studies.8, 9, 13 Few studies that have examined histological subtype among racially diverse populations have found that compared to whites, blacks have a lower percentage of clear cell RCC and a higher percentage of the papillary subtype.14 It remains unclear, however, whether racial differences in histological subtype distributions have contributed to overall racial differences in survival of RCC patients. Compared to white RCC patients, black patients have been shown to have a poorer prognosis despite their generally younger age and earlier tumor stage at diagnosis.15 Some studies attributed the survival disadvantage among black patients, in part, to differential access to surgical treatment,16, 17 whereas others have contended that the racial disparity in survival cannot be explained by access to nephrectomy.18, 19 To better understand racial differences in survival, we examined the survival patterns of RCC patients reported to the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program by demographic characteristics and tumor stage, size, histological subtype, and surgical treatment.
MATERIALS AND METHODS
Patients diagnosed as having RCC (International Classification of Diseases for Oncology, third edition,20 topography codes C64 or C65 and morphology codes 8140-8576) from January 1, 1992, to December 31, 2007, among residents of 12 population-based cancer registries in the SEER program were identified.21 The registries, covering approximately 14% of the US population, are located in the states of Connecticut, Hawaii, Iowa, New Mexico, and Utah, and the metropolitan areas of Atlanta, Detroit, Los Angeles, San Francisco, San Jose, and Seattle, and 10 counties in rural Georgia. The study period was chosen, because data became available for all 12 areas starting in 1992, and the most recent year of data accrual was 2008. To follow SEER practice and to take full advantage of data for all cases, we used 2007 as the cutoff point for inclusion to allow for at least 1 year of survival experience for all patients. Patients with racial origins other than black or white were excluded because of small numbers and the lack of population data for estimating relative survival rates.
During 1992 to 2007, 62,817 kidney cancer cases were reported among white or black residents in the 12 SEER areas. Of these, we sequentially excluded 638 cases reported by death certificate only, 502 cases diagnosed only at autopsy, 5391 cases whose diagnosis was not microscopically confirmed, 7555 cases with tumor histological types other than adenocarcinoma (including 5078 transitional cell carcinomas, most likely arising from the renal pelvis; 947 nephroblastoma [Wilms tumor], the main type of kidney cancer in children; 390 sarcomas and other specified noncarcinomas; 984 poorly specified carcinomas; and 156 malignant neoplasms), and 9321 cases whose cancer was not a first primary. In addition, we excluded 60 cases whose follow-up time was zero, leaving 39,350 cases that met the selection criteria and were actively followed by SEER.
Relative overall survival rates22 up to 5 years after diagnosis were computed using the SEER*Stat software,23 based on the actuarial (life table) method and 1-month intervals of follow-up. Relative survival is defined as the ratio, expressed as a percent, of the proportion of observed survivors in a cohort of cancer patients to the proportion of expected survivors, taking into consideration the survival experience of a general population of the same time period, age, sex, and race.7 Survival was analyzed by subgroups of RCC patients categorized by SEER historic stage (localized, regional, distant, and unstaged), tumor size (< 4.1 cm, 4.1-7.0 cm, > 7.0 cm, and unknown), histological subtype (clear cell [ICD-O-3 morphology codes 8310-2, 8316, 8320], papillary , chromophobe [8270, 8317], other [all other codes], and poorly specified [8140, 8230, 8342, 8440, 8450, 8504, 8575]), and surgical treatment (no surgery, local excision/partial nephrectomy, radical nephrectomy, and other including exploratory/biopsy and unspecified surgical procedure).
During 1992 through 2007, 39,350 RCC patients meeting our inclusion criteria were diagnosed in the SEER-12 registry areas. Of these, 4359 (11.1%) were black and 34,991 (88.9%) were white (Table 1). Reflecting the rising incidence rates, more than 58% of cases were diagnosed during the second half of the study period (2000-2007). The overall 5-year relative survival rate for whites was 72.6% (95% confidence interval [CI] = 72.0, 73.2), compared to 68.0% (95% CI = 66.2, 69.8) for blacks, yielding a white/black survival rate ratio of 1.07 (95% CI = 1.04, 1.10). Survival rates improved over time, increasing from 67.6% in 1992 through 1999 to 76.7% in 2000 through 2007 among whites and from 64.8% to 70.6% among blacks. Improvement in survival generally occurred for patients with cancer of each stage. The white survival advantage over blacks appeared more prominent in the recent than in the earlier time period (data not shown).
Table 1. Five-year relative survival rates among renal cell carcinoma patients diagnosed 1992-2007, followed through 2008, by demographic and clinical characteristics, SEER 12
No. of Patients (%)
No. of Patients (%)
White/Black Rate Ratio
Tumor Size (cm)
4.1 - 7.0
Compared with whites, a greater proportion of blacks (26.3% vs 18.4%) were diagnosed under age 50 and with localized cancer (66.7% vs 61.9%), whereas the distribution of tumor size at diagnosis was similar among blacks and whites. Clear cell RCC was more prevalent among whites (88.6%) than blacks (77.9%), whereas papillary RCC was more common in blacks (12.5%) than whites (4.5%). A higher percentage of blacks than whites (14.5% vs 10.5%) received no surgery. The proportion receiving partial nephrectomy was similar between blacks (13.6%) and whites (12.2%), and the proportion receiving a radical nephrectomy was higher among whites than blacks (73.4% vs 68.5%). Whites generally had a survival advantage over blacks in almost all categories except for the small percentage of patients who received no nephrectomy. The prognosis for these patients was equally poor in both blacks and whites, and was much worse than patients who were treated with partial or radical nephrectomy. Clear cell RCC conferred a poorer prognosis than papillary RCC; however, the extent of white survival advantage (8%-9% higher) over blacks was similar for both histological subtypes. The survival advantage for patients diagnosed at younger age, diagnosed with chromophobe or papillary subtypes, or diagnosed with early stage or small tumors was maintained consistently throughout the 5-year follow-up period (Fig. 1). Furthermore, within each of these categories, whites maintained their survival advantage over blacks.
Even within categories of clinical stage at diagnosis, survival declined with increasing tumor size at diagnosis (data not shown). Likewise, the more favorable prognosis for papillary and chromophobe RCC than clear cell RCC was maintained for cancers within each clinical stage category. The white survival advantage over blacks was seen for all histological subtypes, although the extent of differences was not consistent across all stage categories.
RCC patients who received no surgery were more likely to have distant stage tumors and have lower 5-year relative survival rates (data not shown). The relative survival rates for patients with localized (99.2% for whites and 93.8% for blacks) or regional (90.4% for whites and 68.8% for blacks) cancers following treatment with partial nephrectomy were higher than corresponding patients treated with radical nephrectomy (93.0% for localized and 67.0% for regional cancers among whites, and 87.3% and 60.3% for blacks, respectively). Regardless of tumor clinical stage and surgical treatment, whites consistently had a survival advantage over blacks.
Finally, we examined the survival experience within a group of relatively homogeneous patients, ie, those who were diagnosed with a clear cell RCC at ages 50 to 74 years and were treated with nephrectomy for their cancers (Table 2). As expected, the 5-year relative rates were highest for patients with localized tumors, and within each tumor stage categories, survival rates tended to decrease with increasing tumor size. Despite the similarities in age, treatment, and tumor characteristics, however, white patients generally still have a survival advantage over black patients across almost all stage and tumor size categories.
Table 2. Five-year relative survival rates* for surgically treated clear cell renal cell carcinoma patients aged 50-74 by race, tumor stage, and tumor size at diagnosis, SEER-12, 1992-2007, followed through 2008
Survival rates were estimated only for categories with 25 or more patients;
No category under Unknown clinical stage had 25 or more patients except for whites with unknown size (N=50, Survival rates 62.9% (95% CI: 45.5% - 76.1%)
In our evaluation of nearly 40,000 RCC patients reported to population-based cancer registries across the United States in NCI's SEER program, we found that white patients consistently had a survival advantage over black patients regardless of sex, age, time period of diagnosis, tumor stage and size, histological subtype, or surgical treatment. The more favorable 5-year relative survival rates among whites than blacks existed even among patients with relatively homogeneous demographic and tumor clinical characteristics.
Our findings suggest that compared to their white counterparts, black Americans not only have an increased risk of developing RCC,2 but they also have a poorer prognosis after a tumor is diagnosed. The survival disadvantage among blacks cannot be explained by age, as blacks tended to be diagnosed at a younger age when prognosis is better.24 Nor can the black survival disadvantage be explained by racial differences in histological subtypes of RCC. Consistent with findings from previous studies,14 our study also found that a higher proportion of blacks than whites were diagnosed with papillary RCC, a histological subtype that has been associated with a better prognosis than clear cell RCC.9, 10, 12 Differences in access to health care and delay in tumor diagnosis also appear not to be an explanation, as blacks were more likely to be diagnosed at the localized stage and the black disadvantage in survival was seen across all tumor stages and sizes at diagnosis. However, further examination of whether inadequate access to quality health care might have contributed to the survival differentials is warranted, given that a higher proportion of blacks than whites had a tumor with unknown stage or size. Consistent with findings from previous studies,16, 17, 19 we also found that a slightly higher percentage of black patients than white patients were not treated with nephrectomy, and that these patients had a significantly lower survival than surgically treated patients. In a previous study that examined survival among Medicare patients, a lower percentage with nephrectomy treatment and higher prevalence of hypertension among blacks were found to have largely explained the lower survival among black than white RCC patients.16 Their results, however, are not directly comparable to our results, because the racial difference in patients receiving nephrectomy was much larger in that study of older patients (61.2% blacks and 70.4% whites, respectively) compared to our study of patients of all ages (82.1% blacks and 85.6% whites, respectively). Other sociodemographic factors may also influence surgical decisions. For instance, radical nephrectomy was more commonly performed in younger age groups and married patients.17 Likewise, partial nephrectomy use was found to be significantly greater for patients with private insurance or a prepaid health plan, high income, treated at urban hospitals (particularly teaching hospitals), and who resided in the northeastern United States.25 Although we did not detect an overall difference in the proportion of black and white patients treated with partial or radical nephrectomy, it would be of value to examine whether socioeconomic and demographic variables may influence the use of advanced diagnostic and surgical techniques that could impact patient survival.
Racial differences in the prevalence of comorbid conditions might have contributed to survival disparities among RCC patients, as suggested by Berndt et al.16 Hypertension is known to be more common among blacks26, 27 and poor control of hypertension has been associated with a greater risk of RCC in blacks than whites.28 Compared to whites, progression from prehypertension to hypertension has been shown to be accelerated in blacks,29 and hypertension-related chronic kidney diseases and end-stage renal disease are known to be more common among black Americans.27 A multicenter European study suggested that RCC patients with end-stage renal disease might have more favorable clinical features, such as smaller tumor size and lower stage, and better survival than RCC patients in the general population.30 The presence of these conditions among RCC patients potentially could heighten cancer surveillance and influence their treatment decisions, as well as the course of their recovery following cancer diagnosis and surgery. Furthermore, other tumor and host characteristics, such as more refined subtype classification (eg, papillary type 1 and type 2 RCC),31 tumor biology,32 and genetic predisposition33 have been suggested to influence RCC patient prognosis and may play a role in racial disparity in survival. Our study, however, did not have data on comorbid conditions such as hypertension and chronic renal diseases, or standardized review of tumor pathology and availability of biological samples for analysis of molecular markers that may influence prognosis. The role of these conditions and tumor characteristics is best examined in clinical series or population studies with access to tumor tissues and standardized pathological review, detailed information on comorbid conditions and other RCC risk factors, detailed records on surgery and other therapies, and long-term follow-up of RCC patients.
Our study has the advantage of including a sizable number of RCC patients reported to cancer registries with near completeness in registration and follow-up of cancer patients.7 The large numbers allowed us to examine racial differences in survival among subgroups of patients with different clinical characteristics. The consistency in findings across subgroups of patients, as well as in a subgroup of patients with relatively homogeneous clinical characteristics, lends credence to our observation of a survival advantage of white over black RCC patients. In our analysis, we chose to use relative survival rates for comparisons because this method, using a procedure well-established by the U.S. SEER for estimating mortality in the source populations, takes into account the underlying mortality experience in different population subgroups.7 To the extent that blacks and whites may experience different life events that culminate in different underlying mortality experience, and that these events cannot be adequately adjusted for using standard information collected in most studies, the relative survival method takes this issue into consideration by adjusting for the underlying mortality in each group. We also chose to examine the overall mortality rather than cancer-specific mortality because we previously observed that deaths for a substantial proportion of RCC patients were ascribed to noncancer causes, which decreased with advancing stage at diagnosis and increased with length of survival.34 In addition, a small proportion of kidney cancer cases were recorded as deaths due to other urinary tract cancers or other cancers. Finally, following SEER convention,7 patients diagnosed during 2004 through 2007 contributed 1 to 4 years of information to the survival estimates, as available. Given that survival generally has improved over time, our 5-year rates for patients diagnosed in the recent time period may be somewhat underestimated.
In conclusion, we found that white RCC patients consistently have a survival advantage over black patients, regardless of sex, age, time period of diagnosis, tumor stage or size, histological subtype or surgical treatment.
This work was supported by the Intramural Research Program of the National Institutes of Health.