Prostate-specific antigen (PSA) testing has dramatically changed the composition of prostate cancer (PCa), making it difficult to interpret incidence trends. New methods are needed to examine temporal trends in the incidence of clinically significant PCa and whether trends vary by race.
To conduct an in-depth analysis of incidence trends in clinically significant PCa, defined as cases in which PCa was the underlying cause of death within 10 yr of diagnosis.
Design, setting, and participants
We extracted incident PCa cases during the period 1975–2002 and associated causes of death and survival through 2012 from nine cancer registries in the population-based Surveillance Epidemiology and End Results program database.
Outcome measurements and statistical analysis
We applied joinpoint regression analysis to identify when significant changes in trends occurred and age–period–cohort models to examine longitudinal and cross-sectional trends in the incidence of fatal PCa.
Results and limitations
Among 51 680 fatal PCa cases, incidence increased 1% per year prior to 1992, declined 15% per year from 1992 to 1995, and further declined by 5% per year through 2002. Age-specific incidence rates of fatal disease decreased >2% per year among men aged ≥60 yr, yet rates remained relatively stable among men aged ≤55 yr. Fatal disease rates were >2-fold higher in black men compared with white men, a racial disparity that increased to 4.2-fold among younger men.
The incidence of fatal PCa substantially declined after widespread PSA screening and treatment advances. Nevertheless, rates of fatal disease among younger men have remained relatively stable, suggesting the need for additional attention to early onset PCa, especially among black men. The persistent black-to-white racial disparity observed in fatal PCa underscores the need for greater understanding of the causes of this difference so that strategies can be implemented to eliminate racial disparities.
We assessed how the incidence of ultimately fatal prostate cancer (PCa) changed over time. We found that the incidence of fatal PCa declined by >50% since the introduction of prostate-specific antigen testing and advances in treatment options; however, incidence rates among younger men remained relatively stable, and younger black men exhibited a 4.2-fold higher risk for fatal disease compared with white men.
Keywords: Cancer trends, Cause-specific mortality, Disease progression, Prostate cancer, Prostate-specific antigen.
Prostate cancer (PCa) is the most frequently diagnosed cancer among men in the United States, with 180 890 new cases estimated for 2016 . Despite notable improvements in PCa mortality rates in the United States over the past few decades , it is estimated that 26 120 men (8% of male cancer deaths) will die from this disease in 2016 . Racial disparities in PCa are higher than for any other malignancy, with black men exhibiting a 2.5-fold greater risk of death from PCa compared with white men  and .
There has been a substantial increase in PCa incidence rates in the United States over the past few decades largely due to the increased detection of asymptomatic disease, first through transurethral prostatectomy (TURP) for benign prostatic hyperplasia , followed by widespread prostate-specific antigen (PSA) testing beginning in 1986  and . The subsequent decrease in PCa incidence rates, attributed to depletion of the latent PCa reservoir in the population, rendered the now-familiar “spike” in overall PCa incidence rates .
Accurate interpretation of clinically significant PCa incidence trends has been difficult because of the high prevalence of indolent disease and changes in screening practices. A majority of studies examining PCa trends have focused on the date of death (mortality rate) as opposed to the date of diagnosis  and , which is a more applicable time metric for assessing disease trends during a period of clinical change. In addition, no study in the past decade examined trends using age–period–cohort analysis , likely because of unstable parameters resulting from the PCa incidence rate spike.
To overcome these problems and to provide an in-depth analysis of clinically significant PCa incidence trends and racial disparities, we assessed trends in the incidence of fatal PCa, defined as death from this disease within 10 yr of diagnosis .
2. Materials and methods
We obtained census population estimates, cancer incidence, mortality, and survival data for calendar years 1975 to 2012 from the National Cancer Institute's Surveillance Epidemiology and End Results (SEER) SEER-9 registries, which cover approximately 10% of the US population. We selected incident cases diagnosed with PCa (International Classification of Diseases for Oncology, third edition, code C619) as the first primary malignant cancer from 1975 (the first year all SEER-9 registries were operational) through 2002. Men were excluded if diagnosis was based on autopsy reports or death certificate only (n = 4137), if follow-up time was unknown (n = 890), or if age at diagnosis was missing (n = 15). Men were defined as having fatal PCa when PCa was the underlying cause of death within 10 yr of diagnosis (follow-up through 2012). The 10-yr window was selected a priori based on empirical evidence from SEER data (80% of all PCa-specific deaths were captured) and literature review  and .
We calculated age-standardized incidence rates and age-specific incidence rates per 100 000 person-years. We conducted joinpoint analysis to identify when statistically significant changes in trends occurred and estimated the annual percentage change in rates between joinpoints using weighted least squares . Age–period–cohort models of fatal PCa were conducted to further examine temporal trends by year of diagnosis and to distinguish between influences that occurred in specific time periods for all age groups (period effects) and effects associated with specific birth cohorts (generational effects)  and . For these models, data were classified by 1-yr periods into 40 age groups (45–84 yr), 28 calendar periods (1975–2002), and 66 birth cohorts (1891–1956). We fitted cross-sectional age-specific rates based on the central 1988 calendar year. We calculated black-to-white incidence rate ratios to examine racial disparities by period and age at diagnosis (5-yr weighted average age groups) . Last, we used a quadratic spline to fit instantaneous hazard rates of death from PCa (percentage dying of PCa per year among all PCa cases), stratified by race and year of diagnosis (1975–1988 vs 1989–2002).
In sensitivity analyses, we evaluated shortened (5-yr) and extended (15-yr) follow-up for cause-specific mortality to assess whether competing risks and lead time bias affected trends. We also evaluated variable lead time, for which we extended follow-up to 15 yr in the PSA era. To evaluate the robustness of our models, we (1) included men with metastatic disease in our case definition; (2) limited the case pool to men for whom PCa was the sole cancer diagnosis; (3) expanded the case pool to include all men diagnosed with PCa, regardless of cancer sequence; (4) used predicted 10-yr probabilities of death due to PCa for men who died from other causes; and (5) concatenated the SEER-9, -13, and -18 databases.
Statistical analyses were performed in Matlab version R2014b (MathWorks Inc, Natick, MA, USA). All statistical tests were two-sided, and P values <0.05 were considered statistically significant.
Of the 309 289 men diagnosed with PCa during 1975–2002, fatal PCa accounted for 17% (n = 51 680) of the cases. A total of 222 321 men (72%) with PCa died from any cause during any period of follow-up, and 150 565 of those deaths (68%) occurred within 10 yr of PCa diagnosis. During the 10-yr period following diagnosis, PCa was the most common cause of death (34%), followed by ischemic heart disease (24%) and cerebrovascular diseases (5%).
3.1. Fatal prostate cancer trends
Trends in age-standardized incidence rates of fatal PCa are shown in Figure 1A. Joinpoint regression analysis of trends by year of diagnosis identified two joinpoints, in 1992 and 1995 (Table 1), which indicated that rates of fatal PCa increased 1.0% per year from 1975 to 1992, declined 15.0% per year during the subsequent 3-yr period (1992–1995), and steadily declined by 5.0% per year through 2002. Age-specific rates of fatal PCa by year of diagnosis exhibited differences between older and younger men, such that rates among younger men appeared relatively stable, whereas rates in older men (aged ≥55 yr) revealed a declining trend beginning in the early 1990s (Fig. 1B).
Temporal trends in age-standardized and age-specific incidence rates of fatal prostate cancer for men aged 45–84 yr, 1975–2002, United States (Surveillance Epidemiology and End Results SEER-9 registries). (A) Age-standardized incidence rates of fatal prostate cancer among all men. (B) Age-specific incidence rates of fatal prostate cancer by year of diagnosis (5-yr age groups). Age-standardized rates are plotted on the log scale and are age adjusted to the 2000 US standard population. All age-specific incidence rates are plotted on the log scale.
Annual percentage changes in age-standardized incidence rates of fatal prostate cancer over the entire study period and for subperiods defined from joinpoint analysis by race, ages 45–84 yr, 1975–2002, United States (Surveillance Epidemiology and End Results registry SEER-9)
|Race||Year||Annual percentage change||95% confidence interval|
|All races||All years||−2.44||−3.27 to −1.61|
|1975–1992||1.02||0.72 to 1.32|
|1992–1995||−14.99||−21.86 to −7.51|
|1995–2002||−4.95||−6.02 to −3.88|
|White||All years||−2.55||−3.38 to −1.71|
|1975–1992||0.97||0.61 to 1.33|
|1992–1994||−18.67||−18.68 to −18.66|
|1994–2002||−5.54||−6.00 to −4.47|
|Black||All years||−1.97||−2.90 to −1.02|
|1975–1993||1.45||0.83 to 2.07|
|1993–1995||−19.48||−19.49 to −19.48|
|1995–2002||−5.93||−8.31 to −3.50|
Annual percentage changes were calculated using the weighted least squares method. All annual percentage changes are significantly different from zero, P < 0.05 (two-sided permutation test).
Age–period–cohort models for fatal PCa were successfully fit. Age-specific annual percentage changes were heterogeneous (two-sided Wald test; χ2 = 166.3, df = 40, p < 0.0001) when compared with the overall annual percentage change of −1.77% per year. Rates among younger men (aged ≤52 yr) increased by up to 0.6% per year, whereas rates decreased for older men (aged ≥61 yr) by >2% per year (Supplementary Fig. 1). Period rate ratios demonstrated a significant decline in risk of fatal disease with calendar year (Wald test; χ2 = 1800.11, df = 26, p < 0.0001) (Supplementary Fig. 2).
3.2. Racial disparities in fatal prostate cancer
There were 42 112 fatal cases among white men (81.5%) and 7434 fatal cases among black men (14.4%). Black men had an earlier mean age at diagnosis (69.1 vs 71.1) and slightly shorter mean survival among those who died within 10 yr (44.4 vs 48.4 mo) (Supplementary Table 1).
Temporal trends in rates of fatal PCa among both black and white men (Fig. 2A) displayed no evidence of heterogeneity (Wald test; p = 0.16). Age–period–cohort analysis indicated a period effect in the early 1990s for both black and white men. The black-to-white rate ratios revealed no improvements in the racial disparity of fatal PCa incidence over the past few decades (Fig. 2B), with black men remaining at a consistent 2.2- to 3.0-fold elevated risk of fatal disease. In addition, cross-sectional rates by age at diagnosis revealed greater racial disparity among younger men (aged <55 yr) compared with older men (Fig. 3A) (Wald test; χ2 = 27.6, df = 1, p < 0.001). Rates of fatal PCa were twofold greater among older black men compared with older white men (aged ≥60 yr), a racial difference that increased to 4.2-fold among younger men (aged 45–49 yr) (Fig. 3B).
Temporal trends in age-standardized incidence rates of fatal prostate cancer among black and white men by year of diagnosis and corresponding black-to-white rate ratios for men aged 45–84 yr, 1975–2002, United States (Surveillance Epidemiology and End Results SEER-9 registries). (A) Fitted temporal trends in the incidence of fatal prostate cancer with scatter points for black (red squares) and white (blue circles) men. (B) Black-to-white period-specific rate ratios in the incidence of fatal prostate cancer, adjusted for nonlinear cohort effects and conditioned on age at diagnosis (64 yr). The age-standardized incidence rates among black and white men are plotted on the log scale.
Cross-sectional age-specific incidence rates of fatal prostate cancer among black and white men and corresponding black-to-white rate ratios (5-yr age group weighted averages) for men aged 45–84 yr, 1975–2002, United States (Surveillance Epidemiology and End Results SEER-9 registries). (A) Cross-sectional age-specific incidence rates of fatal prostate cancer among black (red squares) and white (blue circles) men, adjusted for nonlinear cohort effects and conditioned on the 1988 central calendar year of diagnosis. (B) Black-to-white age-specific (5-yr age group weighted averages) rate ratios in the incidence of fatal prostate cancer. The cross-sectional age-specific incidence rates are plotted on the log scale.
We calculated annual hazard rates of PCa-specific death among all men diagnosed with PCa, stratified by race and period (Fig. 4). During the pre-PSA era (1975–1988), the instantaneous hazard of PCa death was significantly higher for black men compared with white men, yet the hazard curves were approximately parallel, reaching their peaks (black: 8.5% per year; white: 5.5% per year) at 1–1.5 yr after diagnosis. In contrast, annual hazard rates of PCa death were greatly reduced during the PSA era, yet black men retained higher instantaneous hazards and reached their peak earlier (1.5 yr after diagnosis) compared with white men (2.5 yr).
Annual hazard rates of death due to prostate cancer as the underlying cause of death (percentage of prostate cancer cases dying each year following diagnosis) stratified by race and calendar period for men aged 45–84 yr, 1975–2002, United States (Surveillance Epidemiology and End Results SEER-9 registries). Hazard rates are plotted for all prostate cancer cases among white and black men. The period 1975–1988 is classified as the pre–prostate-specific antigen (PSA) era and the period 1989–2002 as the PSA era.
In our sensitivity analysis that assessed shorter-term follow-up (5 yr) for determination of cause-specific mortality, trends were virtually unchanged (Supplementary Fig. 4 and 5). This indicates that the trends observed in the main 10-yr analysis were not due to competing risks. Longer term follow-up (15 yr) to define fatal PCa proved almost identical to our main analysis, suggesting that lead time bias during the pre- and post-PSA eras did not substantially affect results. Analyses that included variable lead time (extended follow-up of 15 yr only within the PSA era) and metastatic PCa did not materially change the trends and disparities observed (Supplementary Fig. 6–13). For all remaining sensitivity analyses, trends were also essentially unchanged from the main analyses presented (data not shown).
This study revealed a substantial decline in the incidence of fatal PCa among US men during the period 1975–2002, yet rates among younger men remained relatively stable over the study period. Our study showed that black men had substantially greater risk of fatal PCa than white men in every period and cohort examined, and this racial disparity was magnified among younger men.
Prior studies that examined trends in PCa mortality , , , and , as opposed to incidence, are likely more accurate in describing the changing burden of disease, given large artifactual changes in incidence attributable to diagnostic tests. The debate continues as to what extent declines in mortality rates since 1991 are attributable to period effects such as the advent of PSA testing , since lead time introduced through earlier detection has obscured trends . By using the hard end point of PCa death and a fixed follow-up period for cause-specific (PCa) mortality, we were able to conduct an in-depth analysis of clinically significant incidence trends based on calendar year of diagnosis as the time metric. Our approach supports and extends previous analyses of PCa mortality trends and metastatic or distant stage incidence trends, which show declining rates of approximately 50% since the early 1990s , , and .
Men with PCa are at a much greater risk of dying from a cause other than their malignancy, and the percentage of men who die from PCa decreases with increasing time since diagnosis . The rise and fall of US PCa mortality rates is likely the result of a variety of factors including real changes in disease trends, TURP , PSA testing , advances in treatment modalities , changes in lead time , and accuracy of cause-specific mortality . Although our study cannot tease out the specific contribution of each period effect to temporal changes observed in the incidence of fatal PCa, our comprehensive study of clinically significant disease supports the idea that the majority of the decline in incidence is likely due to a combination of improved detection and treatment advances.
This study shows that black men were consistently at greater risk in every period and cohort examined, and race-specific models of fatal PCa revealed nearly parallel trends in age, period, and cohort effects. Recent studies of PCa mortality in the United States have shown the black-to-white disparity beginning to narrow over the past decade , yet studies prior to 2007 reported that the racial disparity was rising . In our study, in which we assessed the incidence of fatal PCa through 2002, we observed no improvement in the black-to-white disparity over the 28-yr period examined.
The black-to-white disparity in clinically significant PCa is poorly understood. Access to care has been postulated as a contributing cause of the racial disparity, yet black and white men have had approximately equal prevalence of PSA testing  and , and racial differences long preceded the PSA era. However, recent evidence has found that black race and low income are associated with lower rates of aggressive treatment of PCa among men with localized or regional disease . Further research on the impact of differences in treatment by race will aid in providing insight into racial disparities in survival. Genetics and the tumor microenvironment may explain a majority of racial disparities seen in PCa  and . In men of African ancestry, fine mapping of the 8q24 region, which is known to harbor multiple PCa-susceptibility risk variants, has identified three new ancestry-specific risk variants .
In 2016, >10% of incident PCa diagnoses are expected to occur in men aged <55 yr , and men with early onset disease have a different clinical profile and typically a stronger genetic component . To our knowledge, only one study to date has observed the heightened racial disparity among younger men and found a 4.9 times greater risk of advanced or metastatic PCa in younger black men compared with white men (aged 40–49 yr) . Our results support this finding: We observed that the black-to-white racial disparity in fatal PCa increased to 4.2 times among younger men (aged 45–49 yr). In addition, it is notable that rates of fatal PCa among younger men remained relatively stable even after the introduction of PSA testing and may be related to low PSA testing utilization among these age groups . Future studies should strive to provide additional attention to early onset PCa, as rates of fatal disease among younger men do not appear to be declining. Age- and race-specific trends in clinically significant PCa may help shape future screening guidelines.
Strengths of this study include the length of the study period; the quality of the registry data; and the availability of a large number of cases from a representative, regionally diverse population in the United States. In addition, we complemented standard descriptive methods with age–period–cohort models. Limitations include the typical caveats associated with registry-based descriptive studies (missing data, lack of case-specific risk factors) and assumptions inherent to age–period–cohort models (the nonidentifiability issue). Our use of 10 yr of follow-up to determine fatal disease status may not be long enough to completely eliminate lead time bias or nullify calendar period shifts in lead time attributable to changes in diagnostic practices. However, sensitivity analyses using extended and variable follow-up periods provided similar results to our main analysis.
Incidence rates of fatal PCa in the United States have declined by >50% since the early 1990s, coinciding with diagnostic and treatment changes, yet rates among younger men have remained relatively stable. Robust risk-triaging strategies are needed to help further reduce PCa that ultimately results in fatal disease and to limit undue anxiety and overtreatment of men with PCa that will not progress. The racial disparity in fatal PCa persists, and the size of this disparity—particularly among younger men—is a call for us to redouble our efforts to elucidate causes and enact strategies to eliminate the disparity.
Author contributions: Michael B. Cook had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Cook, Kelly.
Acquisition of data: Kelly.
Analysis and interpretation of data: Anderson, Andreotti, Cleary, Cook, Kelly, Rosenberg, Younes.
Drafting of the manuscript: Cook, Kelly.
Critical revision of the manuscript for important intellectual content: Anderson, Andreotti, Cleary, Cook, Kelly, Rosenberg, Younes.
Statistical analysis: Cook, Kelly, Rosenberg.
Obtaining funding: None.
Administrative, technical, or material support: None.
Other (specify): None.
Financial disclosures: Michael B. Cook certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.
Funding/Support and role of the sponsor: This study was supported entirely by the Intramural Research Program of the Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Bethesda, MD. No funding or other financial support was received.
Acknowledgments: This article was presented in poster form at the American Association for Cancer Research 2016 annual meeting (April 20, 2016; New Orleans, LA, USA). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This manuscript is not presently under consideration for publication elsewhere, nor have the contents of this manuscript been copyrighted or published previously.
-  R.L. Siegel, K.D. Miller, A. Jemal. Cancer statistics, 2016. CA Cancer J Clin. 2016;66:7-30
-  A.B. Ryerson, C.R. Eheman, S.F. Altekruse, et al. Annual Report to the Nation on the Status of Cancer, 1975–2012, featuring the increasing incidence of liver cancer. Cancer. 2016;122:1312-1337
-  C. DeSantis, D. Naishadham, A. Jemal. Cancer statistics for African Americans, 2013. CA Cancer J Clin. 2013;63:151-166 Crossref
-  R.M. Merrill, E.J. Feuer, J.L. Warren, N. Schussler, R.A. Stephenson. Role of transurethral resection of the prostate in population-based prostate cancer incidence rates. Am J Epidemiol. 1999;150:848-860 Crossref
-  T.A. Stamey, N. Yang, A.R. Hay, J.E. McNeal, F.S. Freiha, E. Redwine. Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. N Engl J Med. 1987;317:909-916 Crossref
-  H.G. Welch, P.C. Albertsen. Prostate cancer diagnosis and treatment after the introduction of prostate-specific antigen screening: 1986–2005. J Natl Cancer Inst. 2009;101:1325-1329 Crossref
-  M.S. Wachtel, T. Nelius, A.L. Haynes, S. Dahlbeck, W. de Riese. PSA screening and deaths from prostate cancer after diagnosisA population based analysis. Prostate. 2013;73:1365-1369 Crossref
-  R. Etzioni, A. Tsodikov, A. Mariotto, et al. Quantifying the role of PSA screening in the US prostate cancer mortality decline. Cancer Causes Control. 2008;19:175-181 Crossref
-  K.C. Chu, R.E. Tarone, H.P. Freeman. Trends in prostate cancer mortality among black men and white men in the United States. Cancer. 2003;97:1507-1516 Crossref
-  R.T. Vollmer. The dynamics of death in prostate cancer. Am J Clin Pathol. 2012;137:957-962 Crossref
-  M.M. Epstein, G. Edgren, J.R. Rider, L.A. Mucci, H.O. Adami. Temporal trends in cause of death among Swedish and US men with prostate cancer. J Natl Cancer Inst. 2012;104:1335-1342 Crossref
-  H.J. Kim, M.P. Fay, E.J. Feuer, D.N. Midthune. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med. 2000;19:335-351
-  T.R. Holford. The estimation of age, period and cohort effects for vital rates. Biometrics. 1983;39:311-324 Crossref
-  P.S. Rosenberg, W.F. Anderson. Age-period-cohort models in cancer surveillance research: ready for prime time?. Cancer Epidemiol Biomarkers Prev. 2011;20:1263-1268 Crossref
-  R. Etzioni, R. Gulati, A. Tsodikov, et al. The prostate cancer conundrum revisited: treatment changes and prostate cancer mortality declines. Cancer. 2012;118:5955-5963 Crossref
-  G. Draisma, R. Etzioni, A. Tsodikov, et al. Lead time and overdiagnosis in prostate-specific antigen screening: importance of methods and context. J Natl Cancer Inst. 2009;101:374-383 Crossref
-  E.J. Feuer, R.M. Merrill, B.F. Hankey. Cancer surveillance series: interpreting trends in prostate cancer - part II: cause of death misclassification and the recent rise and fall in prostate cancer mortality. J Natl Cancer Inst. 1999;91:1025-1032 Crossref
-  H.G. Welch, D.H. Gorski, P.C. Albertsen. Trends in metastatic breast and prostate cancer - lessons in cancer dynamics. N Engl J Med. 2015;373:1685-1687
-  M.R. Cooperberg, A.J. Vickers, J.M. Broering, P.R. Carroll. Comparative risk-adjusted mortality outcomes after primary surgery, radiotherapy, or androgen-deprivation therapy for localized prostate cancer. Cancer. 2010;116:5226-5234 Crossref
-  C. Zeng, W. Wen, A.K. Morgans, W. Pao, X. Shu, W. Zheng. Disparities by race, age, and sex in the improvement of survival for major cancers: results from the national cancer institute surveillance, epidemiology, and end results (seer) program in the United States, 1990 to 2010. JAMA Oncol. 2015;1:88-96 Crossref
-  A.A. Aizer, T.J. Wilhite, M.H. Chen, et al. Lack of reduction in racial disparities in cancer-specific mortality over a 20-year period. Cancer. 2014;120:1532-1539 Crossref
-  A.B. Mariotto, R. Etzioni, M. Krapcho, E.J. Feuer. Reconstructing PSA testing patterns between black and white men in the US from Medicare claims and the National Health Interview Survey. Cancer. 2007;109:1877-1886 Crossref
-  M.B. Cook, P.S. Rosenberg, F.A. McCarty, et al. Racial disparities in prostate cancer incidence rates by census division in the United States, 1999–2008. Prostate. 2015;75:758-763 Crossref
-  D.R. Ziehr, B.A. Mahal, A.A. Aizer, et al. Income inequality and treatment of African American men with high-risk prostate cancer. Urol Oncol. 2015;33 18.e7–13
-  I.J. Powell, G. Dyson, S. Land, et al. Genes associated with prostate cancer are differentially expressed in African American and European American men. Cancer Epidemiol Biomarkers Prev. 2013;22:891-897 Crossref
-  T.A. Wallace, R.L. Prueitt, M. Yi, et al. Tumor immunobiological differences in prostate cancer between African-American and European-American men. Cancer Res. 2008;68:927-936 Crossref
-  Y. Han, K.A. Rand, D.J. Hazelett, et al. Prostate cancer susceptibility in men of African ancestry at 8q24. J Natl Cancer Inst. 2016;108:djv431
-  C.A. Salinas, A. Tsodikov, M. Ishak-Howard, K.A. Cooney. Prostate cancer in young men: an important clinical entity. Nat Rev Urol. 2014;11:317-323 Crossref
-  I.J. Powell, C.H. Bock, J.J. Ruterbusch, W. Sakr. Evidence supports a faster growth rate and/or earlier transformation to clinically significant prostate cancer in black than in white American men, and influences racial progression and mortality disparity. J Urol. 2010;183:1792-1796
-  M.W. Drazer, D.Z. Huo, M.A. Schonberg, A. Razmaria, S.E. Eggener. Population-based patterns and predictors of prostate-specific antigen screening among older men in the United States. J Clin Oncol. 2011;29:1736-1743 Crossref
a Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA
b Department of Epidemiology and Biostatistics, Milken Institute School of Public Health, The George Washington University, Washington, DC, USA
Corresponding author. Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, 9609 Medical Center Drive, Room 7E-106, MSC 9774, Bethesda MD 20892, USA. Tel. +1 240 276 7298; Fax: +1 240 276 7838.
© 2016 Published by Elsevier B.V.