Global Patterns of Prostate Cancer Incidence, Aggressiveness, and Mortality in Men of African Descent
Prostate cancer (CaP) is the leading cancer among men of African descent in the USA, Caribbean, and Sub-Saharan Africa (SSA). The estimated number of CaP deaths in SSA during 2008 was more than five times that among African Americans and is expected to double in Africa by 2030. We summarize publicly available CaP data and collected data from the men of African descent and Carcinoma of the Prostate (MADCaP) Consortium and the African Caribbean Cancer Consortium (AC3) to evaluate CaP incidence and mortality in men of African descent worldwide. CaP incidence and mortality are highest in men of African descent in the USA and the Caribbean. Tumor stage and grade were highest in SSA. We report a higher proportion of T1 stage prostate tumors in countries with greater percent gross domestic product spent on health care and physicians per 100,000 persons. We also observed that regions with a higher proportion of advanced tumors reported lower mortality rates. This finding suggests that CaP is underdiagnosed and/or underreported in SSA men. Nonetheless, CaP incidence and mortality represent a significant public health problem in men of African descent around the world.
Little is known about the epidemiology of CaP among men in Sub-Saharan Africa (SSA) . However, men of SSA descent around the world appear to suffer disproportionately from CaP compared to men of other races or ethnicities . The International Agency for Research on Cancer (IARC) estimates that CaP is the leading cancer in terms of incidence and mortality in men from Africa and the Caribbean . IARC also estimates that CaP is a growing problem in Africa with approximately 28,006 deaths from CaP in 2010, and approximately 57,048 deaths in 2030 . This represents a 104% increase in the number of CaP deaths in Africa over the next two decades. However, CaP incidence and mortality rates that may be underestimated in SSA and possibly the Caribbean were compared to USA rates due to lack of screening, limited population-based cancer registry data, and underdiagnosis or treatment.
Comparisons of CaP incidence, prevalence, aggressiveness, and mortality in men of SSA descent are limited [5, 6]. Jackson et al.  reported that the age-adjusted incidence rate in a combined West African (Accra and Ibadan) series (36.7 per 1000) was almost equal to the rate in the Washington, DC series (40.6 per 1000). Although comparable prevalence data for Senegal are not available, Gueye et al.  compared clinical characteristics of CaP in African American (AA) and Senegalese men diagnosed between 1997 and 2002. Tumor stage was more advanced in Senegalese men than in AA men. These differences may reflect symptom-based diagnosis in Senegal compared to diagnosis following elevated PSA and/or positive digital rectal examination (DRE) in the USA. A number of papers in African populations have characterized PSA levels, use of screening, and the relationship of PSA with prostate tumor characteristics [9–13]. Similar differences between prostate cancer characteristics among men in New York, Guyana, and Trinidad have been reported with respect to tumor aggressiveness and survival . These data suggest that important differences by geography exist among men of African descent and underscore the need for accurate CaP data for Caribbean, SSA, and AA populations in order to make meaningful comparisons. Insufficient population-based cancer registration in SSA continues to be a barrier to obtaining high quality data.
2.1. Sources of Incidence and Mortality Rates
The primary data used for this report were derived from three sources. First, age-adjusted African-American rates were obtained for 2008 incidence from the 17 registry SEER data set (SEER-17)  and from 2008 mortality from national data  using SEER*Stat software . National estimates of the number of cases were derived by multiplying the age-specific SEER-17 incidence rates by the corresponding national population estimates and summing over all age groups. For presentation of incidence time trends, we used the 13 registry SEER data set (SEER-13)  for the years 1992–2008. The USA data were tabulated using SEER*Stat software . Second, estimated world incidence and mortality rates for the rest of the world were obtained from GLOBOCAN 2008 . As with SEER data, these rates were adjusted according to the 1960 world standard. Third, a complete literature review was undertaken to identify studies reporting CaP rates in men of African descent by searching Medline for the keywords “prostate cancer”, “African”, “Caribbean”, and “Black.” Studies were initially not excluded based on sample size, study design, or methodological quality. To maximize comparability of studies across geography and/or time, studies were excluded if the ascertainment strategy or population base from which the estimates were derived could not be defined; estimates were based on hospital records, death certificates, or other data sources deemed to be potentially unreliable, or if rates were not age-standardized to the 1960 world population. All rates presented here were age-standardized to the 1960 world population to be comparable with other reports of cancer incidence. Data for Africa and the Caribbean from GLOBOCAN 2008 reflect the total population of the country or region being reported, and therefore include men who are of non-African descent. The proportion of men of non-African descent in each country or region is shown in Supplementary Table 1 available online at http://dx.doi.org/10.1155/2013/560857 if known. To limit the inferences to countries where the majority of the population is of African descent, we excluded countries where <50% of individuals were reported to be of African descent (e.g., Puerto Rico and Cuba).
2.2. Studies and Consortia
The men of African descent and Carcinoma of the Prostate (MADCaP) consortium and the African Caribbean Cancer Consortium (AC3) contributed data to this report. The goal of these consortia is to develop multicenter studies that address CaP research at a variety of levels in men of SSA descent in North America, the Caribbean, Europe, and Africa. MADCaP and AC3 studies include epidemiological, biomarker, genetic, risk assessment, and outcomes prediction research. Data regarding prostate tumor characteristics were collected from the following studies: the CAPGenes Prostate Cancer Genetics Studies (CAPGenes) , Fred Hutchinson Cancer Research Center (FHCRC) CaP Studies [20, 21], the Prostate Cancer Risk Assessment Program (PRAP) at Fox Chase Cancer Center , the Flint Men’s Health Study (FMHS) [23, 24], Gene-Environment Interaction in CaP (GECAP) Study at Henry Ford Hospital , Los Angeles County Study (LACS) , CaP Clinical Outcome Study (PC2OS) at the University of Louisville , MD Anderson Cancer Center , the Multiethnic Cohort Study (MEC) , Moffitt Cancer Center Study , NCI Prostate Tissue Study (NCIPTS), University of Pennsylvania Study of Cancer Outcomes, Risk, and Ethnicity (SCORE) , University of Texas San Antonio Center for Biomarkers of Risk for CaP (SABOR), University of Texas Health Science Center at San Antonio [32, 33], San Francisco Bay Area Prostate Cancer Study (SFBAPCS) , United Kingdom Genetic CaP Study (UKGPCS), and the Wake University Consortium including participants from the Johns Hopkins University, Wake Forest University, and Washington University (St. Louis) , the PROGRES (Prostate Genetique Recherche Senegal) study in Dakar, Senegal [8, 36], the Ghana Prostate Health Study (GPHS) in Accra, Ghana , the Tygerberg Hospital study in Cape Town, South Africa , the Guyana Tumor Registry , the Tobago Prostate Cancer Screening Study , and unpublished data from the Bahamas Prostate Cancer Study. At all MADCaP and AC3 centers, data were collected under protocols approved by local IRBs.
2.3. Assessment of CaP Ascertainment
Correlations between percent T1 stage tumors and CaP incidence, mortality, and population statistics on medical care were computed using Spearman rank correlations, using abstracted literature review and tumor stage data from the MADCaP consortium. Sources of percent of gross domestic product spent on health care outside the USA were obtained from WHO World Health Statistics 2010 . Comparable data for the USA were obtained from the Centers for Medicare and Medicaid Services by state . STATA 10 was used for all descriptive and statistical analyses of prostate tumor characteristics, incidence, and mortality trends. All values are based on two-sided tests at . Statistically significant differences from the null hypothesis were inferred at .
3.1. Recent Estimates of International Cancer Incidence and Mortality
We estimated that more than 30,000 cases of CaP were diagnosed among AA men in 2008 (Table 1). The GLOBOCAN estimate for men in all of Africa in 2008 was more than 39,000 CaP cases; the number in SSA would be smaller, but was not estimated. The estimated number among Caribbean men was half of the number in the USA or Africa. CaP was by far the most frequently diagnosed reportable cancer during 2008 in AA, Caribbean, and SSA men, with the age-standardized CaP incidence rates ranging from 159.6 per 100,000 among AA to 71.1 in the Caribbean and 17.5 in Africa. The 28,000 deaths due to CaP in Africa were more than five times the 4,600 among AA and four times the 6,500 among Caribbean men. The mortality rate for CaP exceeded that for any other cancer in Africa (12.5 per 100,000) and the Caribbean (26.3 per 100,000), and followed only lung cancer among AA (22.4 versus 49.1 per 100,000). Whereas the CaP incidence rate was highest among AA men, the CaP mortality rate was highest in Caribbean men. Both the CaP incidence and mortality rates were lower in Africa than in AA or Caribbean men. As a result, the mortality : incidence rate ratios ranged from 0.71 in Africa to 0.41 in the Caribbean and 0.14 in AA.
3.2. International Variation in CaP Incidence
Figure 1 presents the age-standardized CaP incidence rates in men of African descent from SSA, the Caribbean, and the USA SEER program for various years 1990–2008. Supplementary Table 1 presents the incidence rates and references for the information plotted in Figure 1. Supplementary Table 2 presents rates that were not included in Figure 1 because they represent estimates that were not standardized to the 1960 world population, did not reflect a clearly defined reference population, were hospital or clinic-based, or for which the sampling or ascertainment frame was not clearly defined. Many of these rates vary widely, suggesting that they may not be accurate. As shown in Figure 1 (and Supplementary Table 1), incidence rates in Africa historically have been <50 per 100,000, based on data from Uganda and Zimbabwe. Rates in SEER-13 and SEER-17 cancer registries have all exceeded 150 per 100,000, with a pronounced peak at 213.1 during 1993 after the introduction of widespread and adoption of PSA testing and nadir at 163.4 during 2005.
During 2008, CaP incidence internationally was highest in the USA Atlanta SEER registry (198.8 per 100,000) and lowest in Niger (5.1 per 100,000), with substantial variability by geography. Within the SEER-17 cancer registries, after excluding two registries with <16 cases, rates ranged from 134.8 per 100,000 in Iowa to 198.8 per 100,000 in Atlanta. In the Caribbean, rates varied from 51.1 per 100,000 in Jamaica to 173.7 per 100,000 in Martinique. In SSA, incidence rates were generally lower than those in the USA or Caribbean varying from 5.1 per 100,000 in Niger to 59.7 per 100,000 in South Africa.
3.3. International Variation in CaP Aggressiveness
Table 2 reports characteristics of prostate tumors in Africa, the Caribbean, the UK, and the USA based on literature review and data from the MADCaP Consortium. These data represent clinical data from observational (e.g., cohort, case-control, or population-based screening) studies, and therefore may not represent the distribution of traits in the population as a whole. In addition, some studies have preferentially ascertained advanced disease, hospital-based cases, or population-based cases. Some imposed age restrictions on case ascertainment. Therefore, the proportions reported here reflect not only geographic differences but also differences in study design. The proportion of tumors that were T1 stage in AA men ranged from 35–63%, compared with 15–29% in SSA, and 24% in the Caribbean and 34% in the UK. There were significant differences in Gleason score (), tumor stage (), and PSA levels at diagnosis () between USA, UK, and African studies. A significantly greater proportion of tumors in Africa had a high Gleason score or high tumor stage compared with those in the USA or UK. In the UK and USA, the most common Gleason scores are 6 and 7, whereas the distribution in other regions is more uniform, with lower (≤5) and higher (≥8) scores representing a greater proportion of tumors than in the USA. The tumor stage distribution differed substantially by geography, with a larger proportion of stage T3/T4 tumors in Africa than other locations. PSA levels at diagnosis were lowest in the USA and Tobago. There were no statistically significant differences in Gleason score (), tumor stage (), or PSA at diagnosis () among centers within Africa.
3.4. International Variation in CaP-Specific Mortality
Table 1 suggested that the CaP ratio of mortality to incidence in Africa (0.71) is notably higher than that in the Caribbean (0.41) or AA men (0.14). To further explore mortality rates in these regions, Figure 2 presents CaP mortality rates (standardized to the 1960 world population) in men of African descent in the USA, SSA, and the Caribbean. The mortality estimates vary considerably and overlap across regions. In the Caribbean, CaP mortality rates were comparable to or higher than in AA men. For example, the total AA mortality rate was 22.4 per 100,000 compared with estimates as high as 61.7 per 100,000 in Barbados , which was the highest reported in the world. In SSA, estimated mortality rates were generally lower than those in the USA or Caribbean. Mortality estimates in 2008 ranged from a low of 4.4 per 100,000 in Niger to 28.8 per 100,000 in Cote d’Ivoire . Indeed some rates in AA were lower than several in Africa and many in the Caribbean. As shown in Table 2, the proportion with high stage/grade disease at diagnosis is much greater in Africa than in AA.
3.5. Completeness of CaP Ascertainment
Figure 3 presents the relationship of the proportion of tumors diagnosed at T1 stage, based on data from the MADCAP consortium, with characteristics that may be related to CaP detection: CaP incidence, based on cancer registry data (Figure 3(a)), percent of gross domestic national or state product spent on health care (Figure 3(b)), and national or state number of physicians per 10,000 persons (Figure 3(c)). As shown in Figures 3(a)–3(c), there was a strong positive correlation between each of these variables and percent of tumors diagnosed at T1 stage. Figure 3(d) also presents the relationship of T1 stage tumors by CaP mortality in SSA, the Caribbean, and several USA states for which T1 stage data were available. Figure 3(d) suggests that the higher the proportion of T1 stage tumors diagnosed, the higher the mortality rate. Stated differently, mortality rates are lower in SSA and the Caribbean, where T1 stage tumors are least likely to be diagnosed. However, given that the mortality : incidence rate ratio (Table 1) of cancers in Africa (0.71) and the Caribbean (0.41) are substantially higher than in the USA (0.14), these results suggest that there is a substantial underreporting of cases of CaP in at least some parts of SSA and the Caribbean.
Our data suggest that among men of Africa descent, CaP is most common in AA and Caribbean men, and considerably less common in SSA (Figure 1). These data also suggest that CaP is a major cancer in men of African descent throughout the world, and that the currently available incidence and mortality rates may represent an underestimate of the actual CaP incidence and mortality rates in SSA and the Caribbean. Despite the evidence that CaP occurs at high rates in AA men, relatively little data are available regarding the epidemiology of CaP in men of African descent in other locations. Possible explanations for the wide range in CaP incidence and mortality by geography observed here fall into several categories: (1) differences in health care access, diagnosis, and screening; (2) differences in the methodology used to generate rates including completeness of ascertainment and (3) underlying differences in risk due to demographic differences, genetics/biology, lifestyle, or environmental exposures.
4.1. Health Care Access and Screening
PSA is widely used in the USA for detection of CaP. While PSA improves CaP detection beyond digital rectal examination (DRE) , its sensitivity and specificity are suboptimal, and it has not been widely implemented in clinical practice in many parts of the world. PSA testing is responsible for a sharp peak in the incidence of prostate tumors detected in the USA and other countries between 1990–1995, but the additional cases detected during this period were primarily localized disease . Many of these tumors may not have been diagnosed until later, if at all, or until they caused clinically relevant events had PSA testing not been in use . Data regarding screening practices and rates, as well as time trends in mortality, are not widely available for African descent populations outside the USA.
The use of PSA testing impacts the international comparisons made here. While the prevalence of PSA testing is not well documented, it appears that CaP is also commonly detected when PSA testing is performed in Caribbean men . No data exist regarding the prevalence of PSA usage in Africa, but it is believed that PSA testing is not common. Thus, lower rates of CaP in Africa may in part reflect lower probability of CaP detection than in countries like the USA where PSA and DRE screening are more widely used. Compounding the lower use of PSA testing is issues related to access to cancer-related health care, including diagnosis and treatment of disease. There are currently no data on the proportion of CaP that is never diagnosed in a clinical setting in Africa.
4.2. Methodological Issues in Estimating CaP Incidence and Mortality
Limitations in the availability of data in some regions limit our ability to make strong inferences about the actual rates of CaP in some countries and to make valid comparisons of these rates across countries. Direct comparisons of rates globally are limited by a number of methodological issues. Different age standardizations may be used in different studies. All data presented here present age-standardized rates standardized to the 1960 world population to maximize comparability. Limitations in available data also preclude strong inferences and comparisons of international rates. In many countries, incidence and mortality rates do not separate out men of African descent from men of other ethnicities. For example, the available data in the Caribbean and Africa represent all men, regardless of race. Thus, the rates of CaP reported here are meant to reflect those in men of African descent, but in fact they represent the rates for all men in these countries (aside from the countries we have omitted where the proportion of African descent men was low). We have presented data obtained from population-based cancer registries (Supplementary Table 1), but many reports rely on data obtained from hospital-based or other ascertainment strategies that may not reflect CaP cases in the general population (Supplementary Table 2), or may comprise a biased subset of cases. Thus, some reports suggest very low or very high rates of CaP, but it is not clear whether these reports are valid and they may not be useful in making international comparisons. We have also not presented direct comparisons of rates across populations because data were not available to obtain meaningful variance estimates around incidence or mortality rates across populations. However, to the degree that the rates reflect the correct population estimates based on large samples, values may not be required to compare relative rates across populations. Similarly, reports of mortality in Africa are captured by various means, and the accuracy of cause of death reports in addition to completeness of reporting may be inadequate. Therefore, comparisons of USA, Caribbean, and SSA rates may be limited by differences in data reporting across regions.
4.3. Underlying Risk Profile Differences
At least two hypotheses can be proposed to explain differences in CaP risk across populations of African descent. First, it is possible that CaP is underdiagnosed in the less developed world, particularly in SSA, and the rates are in fact higher than is suggested by existing data due to limited tumor detection. This hypothesis is addressed below. Second, it is possible that CaP incidence is truly lower in the developing world than in the developed world. This might imply that SSA populations are less susceptible to CaP due to genetic risk, environmental exposures, or other risk factors that cause CaP. Differences in life expectancy in developed countries versus Africa may influence CaP rates. Differences in underlying risk profiles across populations may also explain in part differences in risk across populations. Recent studies have identified a large number of loci that are involved in CaP susceptibility, and there is growing evidence that the frequency and contribution of these risk alleles to CaP differ across populations. Chang et al.  reported the multicenter MADCaP study of CaP susceptibility using a sample of nearly 8 000 men of African descent in the USA and UK. They reported that the majority of the loci identified as CaP susceptibility loci in White or Asian populations were not replicated in men of African descent. Haiman et al.  reported a genome-wide association study (GWAS) using a sample of over 6 000 African descent CaP cases and controls that overlapped with the sample of Chang. This GWAS also did not replicate most of the previously reported loci identified in European or Asian descent populations . Haiman et al.  also identified a novel locus on chromosome 17 that had not been identified in prior GWAS. The risk variant at this new locus was only identified in African descent populations and was not present in European descent populations. Explanations for the lack of replication across racial groups include the possibility that the underlying genetic etiology of CaP differs across racial groups, and that the frequency of risk alleles in different populations is sufficiently variable that the effects of these variants cannot be detected in some groups. Finally, it is possible that genetic effects are dependent on nongenetic exposures, such that the main effect of a locus differs depending on the environmental context in which it is acting. Given the major differences in environmental exposures, including infection and lifestyle across populations, this is a likely hypothesis that requires additional study.
Even though environment can be hypothesized to play a major role in CaP risk, there have been few consistent exposures identified to date that are clearly associated with CaP risk. The only epidemiological factors that are uncontested are family history of prostate cancer, age, and race. Therefore, it cannot be clearly determined that environmental exposure differences across populations explain different CaP rates given that the role of these exposures on risk remains unclear.
4.4. Is There Evidence for Underascertainment of CaP in Some Populations?
Our data suggest that CaP rates may be underestimated in SSA and possibly the Caribbean. This underascertainment may be due to CaP underdiagnosis, under-ascertainment, or both. It is not possible using the data available here to distinguish these possible explanations for a potential underestimation of CaP. We evaluated the relationship of early stage tumors (i.e., stage T1) with characteristics of the populations in which CaP is diagnosed (Figure 3). Early stage tumors were strongly correlated with higher incidence, greater percent of gross domestic product spent on health care, and greater number of physicians per population. While certainly not conclusive, these data support the hypothesis that countries where greater medical intervention and access to health care are available have higher rates of early stage prostate tumors. This finding is not surprising but is consistent with the hypothesis that CaP incidence rates in SSA are underestimated. However, it is critical to remember that small numbers of cases, usually obtained from clinic- or hospital-based series, may reflect relatively less stable or representative information than those obtained from population-based cancer registries (e.g., SEER) in the developed world.
Evaluation of the variability in CaP rates is severely limited by the availability of appropriate cancer registries in SSA. According to IARC, approximately 38 cancer registries were active in SSA in 2011, but only two of these were fully compliant with IARC standards. One reason often cited for the lack of compliance with IARC standards is that registries may not have access to all newly diagnosed cancer cases in the population and cannot provide pathology confirmation of diagnoses due to limited pathology resources in many countries. Evaluation of the accuracy of cancer incidence rates obtained from population-based cancer registries in Africa has been assessed in Uganda . The evidence from that assessment suggests accurate cancer rates can be obtained in Africa if standardized protocols are followed. Thus, additional population-based cancer registries should be developed using accepted protocols in SSA to obtain accurate estimates of cancer incidence and mortality.
In addition to limitations in the estimates of incidence and mortality from population registries, it is not currently possible to make global comparisons of differences among these rates using available methods. While it is possible to estimate variances and standard errors around the rates of interest when observed counts are available (e.g., SEER data), the standard in the field is to not estimate variances for GLOBOCAN type estimates because they are based around estimated counts, not actual observed counts. While formulae exist that could be used to estimate variants, these are unlikely to be accurate because there are many sources of uncertainty and many assumptions made in the GLOBOCAN estimation process. Ferlay et al.  report that the some necessary assumptions for variance estimation could be quantified, whereas for others they clearly could not be. Thus, it is not possible with current methodology to obtain a single variance estimate that is appropriate for data from all locations. Since it is not acceptable to make comparisons of rates across different locations based around inaccurate variance estimates, we do not estimate variances, nor do we make direct comparisons of differences in rates across populations.
Although we advocate for improved reporting of cancer data to facilitate research and clinical practice, we are aware that there have been concerns about the overdiagnosis of CaP. Therefore, we do not advocate increased use of PSA screening in SSA, but to develop appropriate approaches to CaP screening in SSA that will limit CaP-related morbidity and mortality.
CaP is estimated to be the most common tumor in SSA and Caribbean men. However, accurate estimates of CaP incidence and mortality, particularly in SSA, do not adequately inform appropriate CaP screening and clinical practice. The data presented here do not provide specific guidance about optimal prostate cancer screening or treatment strategies in Africa. However, it is likely that African-specific risk profiles, disease aggressiveness, and health care access mean that screening and treatment strategies may be different in Africa than those in standard practice in North America or Europe. If CaP is actually more common than is currently understood in African descent populations, then CaP incidence rates are likely to be higher and may represent an even greater public health problem than is currently thought. Additional attention to the problem of CaP in Africa is warranted that could include development of epidemiological research and resources (e.g., high quality population-based cancer registries) to improve our understanding of the CaP burden in African populations; epidemiological and genetics research to understand the risk factors acting in Africa; studies to determine appropriate use of PSA testing; and treatment-related research that will lead to reduced CaP morbidity and mortality in the Caribbean and SSA.
The authors have no financial conflicts to declare, other than those listed below. This research was supported by an award from the Landon Foundation. CaPGenes Prostate Cancer Genetics Studies at the University of California, San Francisco: National Institutes of Health Grants CA88164 and CA127298. Flint Men’s Health Study (FMHS): National Institutes of Health Specialized Project of Research Excellence grant in Prostate Cancer Grant P50CA69568. They would like to thank Anna Ray, Dr. Ethan Lange, Kimberly Zuhlke, Joe Washburn, and the University of Michigan CDNA/Microarray Core for their help with this project. Fox Chase Cancer Center (FCCC) Prostate Risk Assessment Program (PRAP): PA Department of Health Grant no. 98-PADMOH-ME-98155, Department of Defense Prostate Cancer Research Program Physician Research Training Award (W81XWH-09-1-0302), and National Institutes of Health Grant CA150079-01. They are grateful to all participants of the Prostate Cancer Risk Assessment Program at Fox Chase Cancer Center. Fred Hutchinson Cancer Research Center (FHCRC) Prostate Cancer Studies: National Institutes of Health Grants CA056678 and CA092579 (to J. L. Stanford), with additional support from the Fred Hutchinson Cancer Research Center (to J. L. Stanford) and the National Human Genome Research Institute (to E. A. Ostrander). Gene-Environment Interaction in Prostate Cancer (GECAP): National Institutes of Health Grant ES011126. The authors wish to thank the GECAP study staff for their help in recruiting cases and controls, data processing, and management: K. Amend, M. Aubuchon, M. Beavers, K. Bohn, J. Broderick, J. Clayton, A. Jolly, J. Mitchell, R. Rose, and D. Thomas. The authors also thank the Medical Genetics Laboratory staff for DNA processing: N. Ballard, M. McDaniel. The Ghana Prostate Health Study (GPHS): The Ghana Prostate Health Study was conducted by the NCI Division of Cancer Epidemiology and Cancer and supported by the National Institutes of Health Intramural Research Funds. They would like to thank Ghana collaborators, Drs. Yao Tettey, Andrew Adjei, Richard Biritwum as well as study coordinators, Evelyn Tay and Vicky Okyne. Guyana Cancer Registry: The authors are gratefulto Joan Kendall and Jackquilin Embleton for datacollection at the Guyana cancer registry. Thanks toDr. Shamdeo Persaud, Chief Medical Officer, Ministryof Health, Guyana for the support of this work. Los Angeles County Study (LACS): Grant 99-00524V-10258 from the Cancer Research Fund, under Interagency Agreement no. 97-12013 (University of California contract no. 98-00924V) with the Department of Health Services Cancer Research Program and by National Institutes of Health Grant CA84979. MD Anderson Cancer Center (MDACC): National Institutes of Health Grant CA68578, Department of Defense Grant DAMD W81XWH-07-1-0645-01, and National Institutes of Health Grant P50-CA140388. Moffitt Cancer Center: Ernest Amankwah, Thomas Sellers, Julio Pow-Sang, Ardeshir Hakam, Hyun Y. Park, Selina Radlein, Maria Rincon, Babu Zachariah. The Moffitt group was supported by National Institutes of Health Grant CA128813. Multiethnic Cohort (MEC) Study: National Institutes of Health Grants CA63464 and CA54281. PROGRES (Prostate Genetique Recherche Senegal) Study at the Hôpital Général de Grand Yoff and Université Cheikh Anta Diop: Drs. Issa Labou, Aston Ndiaye, Lamine Niang, Maguette Sylla Niang, Coumba Toure Kane, Souleymane Mboup, Gabriel Haas, Al Ruenes, and Ms. Nabou Diop for their support in developing prostate cancer research in Senegal. The PROGRES study was supported by the National Institutes of Health Grant CA103359 (to S. M. Gueye). Prostate Cancer Clinical Outcome Study (PC2OS) at the University of Louisville: The authors thank Tiva Van Cleave and Nicole Lavender for data collection and analysis. Rick A. Kittles for generously donating DNA samples collected from men of African descent. This study was supported in part by the National Institutes of Health Grant CA128028 and the James Graham Brown Cancer Center and the Bucks for Brains “Our Highest Potential” in Cancer Research Endowment. San Francisco Bay Area Prostate Cancer Study (SFBAPCS): Grant 99-00527V-10182 from the California Cancer Research Fund. The Tobago Prostate Survey: This study was supported, in part, by funding or in-kind services from the Division of Health and Social Services, Tobago House of Assembly, the University of Pittsburgh Cancer Institute, the University of Pittsburgh Department of Urology volunteer surgery teams who have performed radical prostatectomies in Tobago, contract DAMD 17-99-1-9015, U.S. Department of Defense, and National Institutes of Health Grant CA84950. UK Genetic Prostate Cancer Study (UKGPCS): This work was supported by Cancer Research UK Grant C5047/A7357. They would also like to thank the following for funding support: The Institute of Cancer Research and The Everyman Campaign, The Prostate Cancer Research Foundation, Prostate Research Campaign UK, The National Cancer Research Network UK, The National Cancer Research Institute (NCRI) UK. The authors acknowledge NHS funding to the NIHR Biomedical Research Centre at The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust. They acknowledge the NCRN nurses and Consultants for their work in the UKGPCS study. A list of the study team and collaborators is available on http://www.icr.ac.uk/ukgpcs. University of Pennsylvania Study of Clinical Outcomes, Risk, and Ethnicity (SCORE): National Institutes of Health Grants CA085074 and P50-CA105641 and a Grant from the Commonwealth of Pennsylvania (to T. R. Rebbeck). The authors wish to thank Drs. D. Goldmann, W. Greer, G. W. Crooks, D. A. Horowitz, D. Farhadi, M.D. Cirigliano, M. Rusk, V. Weil, S. J. Gluckman, C. Bridges, M. L. Walker, and C. Guerra for their invaluable assistance in ascertaining study participants. University of Texas San Antonio Center for Biomarkers of Risk for Prostate Cancer (SABOR): SABOR is a Clinical and Epidemiologic Center of the Early Detection Research Network of the National Cancer Institute, supported by National Institutes of Health Grant CA86402 and Cancer Center Support Grant for the Cancer Therapy and Research Center P30 CA054174. Prevalence samples obtained through support from the American Cancer Society (TURSG-03-152-01-CCE). The authors wish to thank Dr. Joke Beuten for assistance with these genetic studies. Wake Forest Consortium: The work at Washington University (St. Louis) was supported by National Institutes of Health-R01CA112028, the St. Louis Men’s Group Against Cancer and the Par For the Cure Foundation.
Supplementary Table 1 presents the incidence rates and references for the information plotted in Figure 1. Supplementary Table 2 presents rates that were not included in Figure 1 because they represent estimates that were not standardized to the 1960 world population, did not reflect a clearly defined reference population, were hospital or clinic-based, or for which the sampling or ascertainment frame was not clearly defined.
P. Boyle and B. Levin, World Cancer Report 2008, IARC Press, Lyon, France, 2008.
F. T. Odedina, T. O. Akinremi, F. Chinegwundoh et al., “Prostate cancer disparities in black men of African descent: a comparative literature review of prostate cancer burden among black men in the United States, Caribbean, United Kingdom, and West Africa,” Infectious Agents and Cancer, vol. 4, supplement 1, article S2, 2009.View at: Publisher Site | Google Scholar
J. Ferlay, H. R. Shin, F. Bray et al., “GLOBOCAN 2008,” Cancer Incidence and Mortality Worldwide: IARC CancerBase 10, International Agency for Research on Cancer, Lyon, France, 2010.View at: Google Scholar
D. Parkin, S. L. Whelan, J. Ferlay et al., Cancer in Africa: Epidemiology and Prevention, IARC Press, Lyon, France, 2003.
M. A. Jackson, J. Kovi, M. Y. Heshmat et al., “Characterization of prostatic carcinoma among blacks: a comparison between a low-incidence area, Ibadan, Nigeria, and a high-incidence area, Washington, DC,” The Prostate, vol. 1, no. 2, pp. 185–205, 1980.View at: Google Scholar
F. M. Abbiyesuku, O. B. Shittu, O. O. Oduwole, and B. O. Osotimehin, “Prostate specific antigen in the Nigerian African,” African Journal of Medicine and Medical Sciences, vol. 29, no. 2, pp. 97–100, 2000.View at: Google Scholar
A. A. Ajape, A. Babata, and O. O. Abiola, “Knowledge of prostate cancer screening among native African urban population in Nigeria,” Nigerian Quarterly Journal of Hospital Medicine, vol. 19, no. 3, pp. 145–147, 2009.View at: Google Scholar
D. Iya, S. Chanchani, J. Belmonte, D. Morris, R. H. Glew, and D. J. A. van der Jagt, “Prostate specific antigen in Africans: a study in Nigerian men,” Nigerian Journal of Surgical Research, vol. 5, no. 3, pp. 114–119, 2003.View at: Google Scholar
C. A. Okolo, O. M. Akinosun, O. B. Shittu et al., “Correlation of serum PSA and gleason score in nigerian men with prostate cancer,” African Journal of Urology, vol. 14, no. 1, pp. 15–22, 2008.View at: Google Scholar
B. Mutetwa, E. Taioli, A. Attong-Rogers, P. Layne, V. Roach, and C. Ragin, “Prostate cancer characteristics and survival in males of african ancestry according to place of birth: data from Brooklyn-New York, Guyana, Tobago and Trinidad,” Prostate, vol. 70, no. 10, pp. 1102–1109, 2010.View at: Publisher Site | Google Scholar
E. Surveillance and End Results (SEER) Program, SEER∗Stat Database: Incidence-SEER 17 Regs Research Data + Hurricane Katrina Impacted Louisiana Cases, Nov 2010 Sub (2000–2008), Linked To County Attributes-Total U.S., 1969–2009 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch: Bethesda, MD.
E. Surveillance and End Results (SEER) Program, SEER∗Stat Database: Mortality-All COD, Aggregated With State, Total U.S. (1990–2007), 2010, National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch: Bethesda, MD.
S.R. Program, SEER∗Stat software http://www.seer.cancer.gov/seerstat/, 2011, National Cancer Institute: Bethesda, MD.
E. Surveillance and End Results (SEER) Program, SEER∗Stat Database: Incidence-SEER 13 Regs Research Data, Nov 2010 Sub (1992–2008)- Linked To County Attributes-Total U.S., 1969–2009 Counties, 2011, National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch: Bethesda, MD.
J. L. Stanford, K. G. Wicklund, B. McKnight, J. R. Daling, and M. K. Brawer, “Vasectomy and risk of prostate cancer,” Cancer Epidemiology Biomarkers and Prevention, vol. 8, no. 10, pp. 881–886, 1999.View at: Google Scholar
L. N. Kolonel, B. E. Henderson, J. H. Hankin et al., “A multiethnic cohort in Hawaii and Los Angeles: baseline characteristics,” American Journal of Epidemiology, vol. 151, no. 4, pp. 346–357, 2000.View at: Google Scholar
T. R. Rebbeck, J. M. Jaffe, A. H. Walker et al., “Modification of clinical presentation of prostate tumors by a novel genetic variant in CYP3A4. [see comment][erratum appears in J Natl Cancer Inst 1999 Jun 16,91(12):1082],” Journal of the National Cancer Institute, vol. 90, no. 16, pp. 1225–1229, 1998.View at: Google Scholar
J. Beuten, J. A. L. Gelfond, J. L. Franke et al., “Single and multigenic analysis of the association between variants in 12 steroid hormone metabolism genes and risk of prostate cancer,” Cancer Epidemiology Biomarkers and Prevention, vol. 18, no. 6, pp. 1869–1880, 2009.View at: Publisher Site | Google Scholar
M. Jalloh, C. Zeigler-Johnson, M. Sylla-Niang et al., “A study of PSA values in an unselected sample of Senegalese men,” The Canadian journal of urology, vol. 15, no. 1, pp. 3883–3885, 2008.View at: Google Scholar
A. P. Chokkalingam, E. D. Yeboah, A. Demarzo et al., “Prevalence of BPH and lower urinary tract symptoms in West Africans,” Prostate Cancer Prostatic Diseases, vol. 15, no. 2, pp. 170–176, 2012.View at: Google Scholar
C. H. Bunker, A. L. Patrick, B. R. Konety et al., “High prevalence of screening-detected prostate cancer among Afro-Caribbeans: the Tobago prostate cancer survey,” Cancer Epidemiology Biomarkers and Prevention, vol. 11, no. 8, pp. 726–729, 2002.View at: Google Scholar
K. C. M. Coard and D. H. A. Skeete, “A 6-year analysis of the clinicopathological profile of patients with prostate cancer at the university hospital of the West Indies, Jamaica,” British Journal of Urology International, vol. 103, no. 11, pp. 1482–1486, 2009.View at: Publisher Site | Google Scholar
World Health Organization, WHO World Health Statistics 2010, http://www.who.int/whosis/whostat/EN_WHS10_Full.pdf), 2010.
Center for Medicare Medicaid Services, http://www.cms.gov, 2010.
R. Etzioni, D. F. Penson, J. M. Legler et al., “Overdiagnosis due to prostate-specific antigen screening: lessons from U.S. prostate cancer incidence trends,” Journal of the National Cancer Institute, vol. 94, no. 13, pp. 981–990, 2002.View at: Google Scholar
B. L. Chang, E. Spangler, S. Gallagher et al., “Validation of genome-wide prostate cancer associations in men of African descent,” Cancer Epidemiol Biomarkers Preventation, vol. 20, no. 1, pp. 23–32, 2011.View at: Google Scholar
C. A. Haiman, G. K. Chen, W. J. Blot et al., “Genome-wide association study of prostate cancer in men of African ancestry identifies a susceptibility locus at 17q21,” Nature Genetics, vol. 43, no. 6, pp. 570–573, 2011.View at: Google Scholar
J. Ferlay, D. Forman, C. D. Mathers et al., “Breast and cervical cancer in 187 countries between 1980 and 2010,” The Lancet, vol. 379, no. 9824, pp. 1390–1391, 2012.View at: Google Scholar