Disparities in prostate cancer incidence and mortality rates: Solvable or not?

Abstract

Prostate cancer (PCa) is recognized as a disease possessing not only great variation in its geographic and racial distribution but also tremendous variation in its potential to cause morbidity and death and it, therefore, ought not to be considered a homogenous disease entity. Morbidity and death from PCa are disproportionately higher in men of African ancestry (MAA) who are generally observed to have more aggressive disease and worse outcomes following treatment compared to men of European ancestry (MEA). The higher rates of PCa among MAA relative to MEA appear to be multifactorial and related to inherent differences in biological aggressiveness; a continued lack of awareness of the disease and methods of prevention; a lower prevalence of screen-detected PCa; comparatively lower access to quality healthcare as well as systemic and institutionalized disparities in the administration of optimal care to MAA in developed countries such as the United States of America where high-quality care is available. Even when access to quality healthcare is assured in equal access settings, it appears that MAA still have worse outcomes after PCa treatment stage-for-stage and grade-for-grade compared to MEA, suggesting that, inherent racial, ethnic and biological differences are paramount in predicting poor outcomes. This review has explored the different contributing factors to the current disparities in PCa incidence and mortality rates with emphasis on the incongruence in how research has been conducted in understanding the disease towards developing therapies.

1. Introduction

The paradigm of cancer disparities is well established¹ and encapsulates differences in cancer outcomes such as, genetic and biological factors, health care access and socioeconomic factors among others (FIG 1) across population groups. It is reported by the CDC’s Office of Minority Health Equity² that while life expectancy and overall health have improved for most Americans in recent years from cancer, this trend is not experienced by all Americans. Furthermore, in a review by³, African Americans irrespective of gender are reported to face considerable disparity in cancer incidence and mortality rates as African American men have a 25% higher incidence and 43% higher mortality rate for all cancers combined in comparison to African American women^{4, 5}.

Figure 1:

Showing risk factors associated with cancer disparities.

This review will focus on prostate cancer (PCa), one of the most diagnosed malignancies among men globally. PCa is the leading cause of cancer deaths in men worldwide. In general MAA carry the highest burden of PCa worldwide and the disease appears to be more aggessive in comparison to other racial groups leading to higher mortality rates. The disease is the second leading cause of cancer-related deaths in the USA⁶ while the Caribbean has the highest age-standardized PCa mortality in the world⁷. The quest for seeking to make sense of and addressing the current disparity in PCa incidence and mortality rates among men of African ancestry (MAA) cannot be done in isolation of a comprehensive examination of this variable disease for which diagnosis and prognosis are not that lucid. Indeed, tackling the solution of minimizing such a disparity has to engage a concerted effort from public and private sectors, having shown that socioeconomic status (SES) through modifiable risk factors can contribute to a patient succumbing to PCa mortality. Notwithstanding, this review not only highlights the disparity of PCa incidence and mortality rates but the disparity in research geared to understanding the disease towards developing therapies for MAA compared to men of European ancestry (MEA). Fairly recent data indicates a decline in PCa incidence, yet, the overall PCa-related mortality continues to rise among MAA⁸.

2. Prostate Cancer

2.1. The disease

Clinical suspicion of the presence of PCa can be assessed by the detection of an abnormal feeling prostate during the digital rectal examination (DRE) and/or abnormal prostate specific antigen (PSA). PCa is typically diagnosed with the aid of a trans-rectal ultrasound-guided prostate biopsy. Standard prostate biopsy protocols involve pre-procedural antibiotics that are known to penetrate prostatic tissue, given to reduce the likelihood of acute bacterial prostatitis caused by the inevitable inoculation of coliforms (gram negative bacteria) into the prostate during trans-rectal biopsy.

The dilemma in evaluating and treating patients with PCa has always been to accurately and precisely risk stratify the patient with respect to the potential for disease progression and death so that a patient who is destined to run a relatively innocuous course and can be left alone while aggressively treating those fated to die from the disease thereby hopefully averting this outcome. It is now recognized that significant harms have been done by the over-detection and over-treatment of men with innocuous forms of PCa over the last several decades^{9, 10}. This has been caused by the indiscriminate application of prostate-specific antigen (PSA)-based screening to populations of men at low risk of death from PCa. This prompted the United States Preventive Services Task Force (USPSTF) in 2012 to recommend against the use of PSA-based screening in all men¹¹. In making this recommendation the USPSTF failed to employ a more nuanced approach which took into consideration the disproportionately higher risk of death of MAA from PCa. MAA possess a greater likelihood of deriving more benefits than harms from PSA-based screening when compared to men of European ancestry (MEA).

Being of African ancestry is one of only three established risk factors for PCa, the other two being a family history and advancing age. PCa mortality is highest in the Caribbean region where most countries are comprised predominantly of persons of African ancestry^{12, 13}. African-American men are almost 2.5 times as likely to die from PCa than their Caucasian-American counterparts¹⁴. MAA are disproportionately at a higher risk of death from PCa compared to their Caucasian counterparts- stage for stage and this is related to not only differences in biologic aggressiveness but also to systematic under-treatment resulting from institutionalized racism in the approach to treating PCa¹⁵. Significant barriers exist to accessing quality care, continued unawareness of the disease and its risk factors, and unequal distribution of risk factors for poor outcomes such as obesity, cigarette smoking and sedentary lifestyle^{15, 16}.

We now recognize that PSA-based screening for PCa is an effective and efficacious method of preventing death from PCa, but it ought to be applied selectively to populations of men at high risk of morbidity and death from the disease who are most likely to derive maximum benefit from treatment of screen-detected cancers while minimizing harms. Populations of MAA are inherently at highest risk of death from PCa and therefore theoretically stand to benefit the most from screening. It is prudent to focus on preventive approaches on this high-risk group. While primary prevention in terms of modification of diet and lifestyle may have some merit, it really is in secondary prevention or PSA-based screening that an immediate impact on PCa mortality will be seen.

Men belonging to populations characterized by high risk of death from PCa such as Afro-Caribbean populations are recommended to start PSA-based screening at age 40 years. In Jamaica, for instance, the Jamaica Urological Society in partnership with the Jamaica Cancer Society recommends PSA-based screening for PCa beginning at age 40 years and the latter entity actively screens men who walk into its office¹³. Two thirds of urologists belonging to the Caribbean Urological Association support PSA-based screening in Afro-Caribbean men and recommend that Caribbean-specific guidelines should be drafted¹⁷. In Afro-Caribbean and African countries, far too many men present with features of advanced or metastatic PCa, an incurable stage of the disease which contributes to the high mortality rate in the Caribbean and sub-Saharan Africa. Features of locally-advanced or metastatic PCa include lower urinary tract symptoms (LUTS), urinary retention, haematuria, lower limb oedema, anorexia, weight loss and bony pain, usually arising from the lower back or pelvis.

Late presentation of the disease may be related to ignorance surrounding the disease, particularly its asymptomatic nature when organ-confined; financial constraints; lack of insurance coverage; fear of the DRE; fear of a diagnosis of cancer or the potential adverse functional outcomes resulting from treatment such as urinary incontinence and erectile dysfunction; a false perception of manhood, fatalism surrounding the disease and its treatment, a false perception that men don’t die from the disease but with it, lack of specialists, and a mal-distribution of the availability of specialist urological care in which urologists concentrate in the cities but are not present in rural areas. In countries like Jamaica with very mountainous terrain, geographical barriers to accessing care may also play a role.

PCa is recognized as a disease possessing great variation in its potential to cause morbidity and death and therefore it cannot be thought of as a homogenous disease entity. This wide variation in the natural history or phenotypic behaviour of PCa is partially predicated on underlying differences in genotype which confers its biologic aggressive potential. Genotypic differences in PCa are correlated with race and it has inconsistently been found that MAA have poorer outcomes when stage and grade of the disease as well as socioeconomic factors are controlled, suggesting that the genotypes of PCa in MAA are associated with worse outcomes¹⁸. A further challenge is that a patient’s underlying genotype is not static but may be modified epigenetically or because of genetic instability as the disease progresses¹⁹.

2.2. The disease in men of African ancestry

With the current advances in technology for early detection and treatment, there has been a steady decline in PCa incidence as well as mortality among MAA that is particularly evident in the United States²⁰. In contrast, studies show that PCa incidence is increasing for MAA in Africa and the Caribbean^21–24. Disparities in incidence and mortality is most evident between Black and White men^{20, 22, 25}. In the USA, for example, despite the declining rates, PCa incidence and mortality rates have consistently remained highest among MAA in contrast to other racial groups^{20, 26}. Similar disparities have also been observed between Black and White men in Africa²².

Populations in Asia and Europe were reported to have lower PCa incidence and mortality rates in comparison to MAA populations with the exception of Africa where incidence rates were lower than expected, possibly due to under-reporting^{20, 26, 27} (Table 1).

Table 1:

Incidence and mortality rates comparing MAA populations and other Populations

MAD Populations	Age-adjusted rates per 100,000, ASR (World)
	Incidence	Mortality
Africa	26.6	14.6
Caribbean	64.2	25.4
USA - MAD	145.9	23.2
Other Populations	-	-
Europe	62.1	11.3
Asia	11.5	4.5

A review of the incidence and mortality rates in PCa among MAA populations worldwide also reveal that there are similarities as well as differences between and within geographical regions of Africa, the Caribbean and the USA. PCa incidence for MAA is highest in the USA and lowest in Africa while mortality rates are highest in the Caribbean and lowest in Africa (Table 1). Within the USA, variations of PCa incidence and mortality for MAA have been reported, whereby highest PCa incidence was observed in Delaware (ASR (W): 183.1 per 100,000) and lowest in South Dakota (ASR (W): 58.1 per 100,000) and PCa mortality was highest in Mississippi (ASR (W): 29.5 per 100,000) and lowest in Hawaii (ASR (W): 11.5 per 100,000)²⁰. Within the African continent, geographical variations in PCa incidence were also observed with the highest incidence reported in South Africa (ASR (W), 68.0 per 100,000) and lowest incidence in Niger (ASR (W), 4.4 per 100,000,)²⁷. PCa mortality rates, were highest in Benin (ASR (W), 36.3 per 100,000) and lowest in Niger (ASR (W), 3.4 per 100,000)²⁶. Similarly, in Caribbean countries, the French territory Guadeloupe was reported to have the highest PCa incidence in the region (ASR (W), 189.1 per 100,000) and the lowest PCa incidence was reported in Guyana (ASR (W), 39.3 per 100,0000)²⁷. Mortality rates in the Caribbean region also varied geographically with Barbados having the highest mortality rates (ASR (W), 48.0 per 100,000) and Puerto Rico having the lowest mortality rates (ASR (W), 12.3 per 100,000)²⁷.

Mortality:Incidence rate ratios (MIR) as seen in FIG 2 provide another perspective of PCa disparities. The majority of countries in Africa and two Caribbean countries, Haiti and Jamaica were reported to have death rates that are more than half or almost equal to that of PCa incidence rates (MIR 0.52–0.79). This suggests that in these countries the rate of death may be similar to the rate of PCa incidence. In contrast, the rates of PCa deaths are half or less than half the rates of PCa incidence in the majority of Caribbean countries, Asia, US, Canada, Europe and some countries in Africa (MIR 0.50–0.10). Developed countries such as Europe, USA and Canada are reported to have mortality:incidence rate ratios less than 0.25. This suggests that patients with PCa are more likely to survive in these countries. Similarly, three French territories (LeReunion, Martinique and Guadeloupe) and one US State (Puerto Rico) although geographically located in developing regions of the Caribbean and Africa also have mortality:incidence rate ratios less than 0.25. These countries have health systems that are similar to their mainland France or the USA and suggest that improvement in health systems infrastructure may contribute to improving disparities in PCa survival.

Figure 2:

Country-specific Mortality:Incidence Rate Ratios for PCa

The complexity of the geographical patterns of PCa incidence and mortality rates between the USA, Caribbean and Africa, within these regions and also between specific countries in these regions may be explained by multiple factors: including differences in the socioeconomic status; the existence, access or utilization of PCa screening or early detection programs; lack of or limited cancer registration programs and/or differences in health care systems infrastructure. It is also plausible that these factors alone may not fully explain the observed varied geographical patterns of incidence or mortality rates and that behavioral/psychosocial and biological contributors may also be evident and are discussed in subsequent sections of this review.

3. Causes for prostate cancer disparities

3.1. Socio-economic status

While there is increasing evidence that genetic factors may account for the disparities in PCa incidence and mortality rates, socio-economic status (SES) also play a role^{16, 28}. There is no geographical standardized methodology for measuring SES but it is believed that the socio-economic status of an individual, family or community relative to others can be utilized as bellwethers for SES in general.

Other factors such as the timely utilization of diagnostic screening kits, the PSA also provide avenues for assessing potential PCa disparities. In a review by Liu²⁹ it was found that there was greater PSA screening among MEA than MAA in a low SES. Furthermore, it is believed that strategies for PSA screening may benefit from tailoring to the patient’s SES. Interestingly, according to the same review by Liu²⁹, prior to the popularity of the PSA in 1987, there were no reported associations of PCa incidence and SES among varied ethnic groups, but post 1987, PSA screening was more prevalent among the higher SES groups as a result of their access to health care. They also reported that prior to the implementation of PSA based screening, men of a high SES were more likely to engage in healthy lifestyle behaviours. Overall, improved accessibility to health care is linked to men of a higher SES who are more likely to screen for PCa earlier than men from a lower SES which usually results in the latter being more susceptible to impermanence. Self-reported PSA tests were also found to be more common among the highly educated³⁰ which resulted in a higher reported incidence rate of 15–19% of PCa in men having at least a college education compared to men with a high school education or below. Low literacy rates and language barriers in the USA have been associated with advanced PCa and increased mortality rates. Generally, MAA fall in the lower socio-economic quartile in which they experience more advanced and aggressive types of PCa when diagnosed at an earlier age compared to their MEA counterparts.

The PRostate Cancer in Ethnic SubgroupS (PROCESS) study in the United Kingdom investigated whether accessibility to healthcare may be a reason for the correlation observed between race and PCa. Reporting on SES, MAA in the cohort lived in less affluent areas and were inclined to be manual labourers. It was also found that knowledge of PCa among MAA and MEA were commensurate. MEA were aware that age was a risk factor for them while MAA were aware that they had a greater disposition to developing the disease³¹. Highlighting behavioural attitudes between the two groups, although the main source of referral by all men in the study was the general practitioner, it was found that emergency rooms and other departments in the hospital were more points of referral for MAA than their MEA counterparts. The research group concluded that there was no evidence of inequality in accessing healthcare for MAA in comparison to MEA yet PCa incidence rates were higher in MAA.

More research and data are needed to conclusively show whether SES directly accounts for the disparities observed for PCa among MAA compared to other ethnic groups. Suggestions are that SES is however, linked to other factors such as diet, comorbidity and genetics which may interplay causing the exacerbated disparities being observed. Modifiable risk fators such as diet, smoking, alcohol consumption, physical activity, medication and infections have been attributed to PCa in men of a low SES²⁸. MAA in the USA appear to be at an increased risk for PCa based on their meat consumption^32–34. As well, MAA are less inclined to receive more aggressive treatment than MEA¹⁶ and so differences in treatment may therefore account for the disparity in the survival rates. These factors are significant barriers in obtaining similar outcomes for MAA with PCa. Therefore, efforts to improve socio-economic factors together with targeted biological approaches may result in the reduction of racial disparities for PCa to be achievable.

3.2. Transatlantic Trade in Africans

The severe population bottleneck caused by the transatlantic trade in Africans may have caused an unnatural selection by favouring survival in persons who were more testosterone responsive. Based on the conditions under which the slaves were transported, slaves possessing greater oxygen-carrying capacity, lean body mass facilitating heat exchange, thick skin, greater sebum production and harder bones (greater bone mineral density), phenotypes all related to testosterone, would have more readily survived the life-threatening journey³⁵. This would cause a population shift towards greater testosterone responsiveness in Africans in the New World compared to West Africans.

The enhanced fitness due to increased testosterone responsiveness would have been consolidated in the New World by the forced interbreeding of slaves from different regions in Africa. Contemporary attributes of African descendants in the New World such as high rates of PCa incidence and mortality as well as sprinting prowess³⁶ are partly attributable to this population shift in testosterone responsiveness³⁵. Examples include the three-fold higher incidence and mortality of PCa in Afro-Trinidadians compared to Indo-Trinidadians^{37, 38} and, high bone density being associated with PCa in Afro-Caribbean men³⁹.

3.3. Research related barriers

3.3.1. In vitro (cell lines)

It took scientists approximately 44 years from the first documented successful growth of animal cells outside the host organism⁴⁰ to developing the first human cancer cell line, HeLa⁴¹. The presence of the HeLa cells paved the way for all other animal cancer and normal cell lines including noteworthy PCa cancer cell lines, LNCaP, DU-145 and PC-3^42–44. The presence of these three PCa cell lines were the spring board for an additional 197 sub-cell lines^{45, 46} of which 95% of them are from MEA as observed from the American Type Culture Collection website. Furthermore, 97% of the PCa cell lines available to researchers globally from main suppliers, ATCC, Sigma and European Collection of Authenticated Cell Cultures (ECACC) are from MEA⁴⁷.

As mentioned, PCa varies from patient to patient, hence no PCa case is exactly alike⁴⁸. Given the limited number of available PCa cell lines from MAA, this creates a challenge when using the few available MAA PCa cell lines for research, as there is no such thing as the “perfect cell line.” No single cell line is capable of adequately representing all the characteristics of PCa. Cell lines in general display observable genetic and morphological changes from the host tumor, based on the number of passages they undergo⁴⁹. Importantly, most cell lines have been grown in vitro for decades which has resulted in genetic drifts and changes at the molecular levels over long periods of time⁵⁰. Furthermore, cancer cell lines have been known to exhibit a greater growth rate than that of the tumors from which they originate. Through various growth conditions such as monolayer culture as opposed to 3-D tumor environments or microenvironment provided by the extra cellular matrix, cells change over time⁵¹. Screening mechanisms might be modelled to identify treatment options based on the characterization specific to a cell line used, but due to molecular changes, results may reflect the genetic changes experienced by the cell line at the point in time⁴⁹.

As PCa cell lines play a role in drug development, the presence of the 3 popular PCa cell lines (LnCap, DU-145 and PC-3) contributed to the advancement of docetaxel as an anti-PCa drug⁵². Indeed, clinical research has showed that docetaxel in conjunction with prednisone improves overall survivorship of men with PCa⁵³ and even though the ethnic backgrounds of the subjects were not revealed, research shows that over 90% of patients enrolled in clinical studies overall for PCa research are Caucasian in origin¹⁸. The correlation between the ethnic backgrounds of PCa cell lines used in anticancer research and their effectiveness in one race over others needs further delineating. Notwithstanding, over the years in vitro, models have focused on PCa cell lines from MEA towards identifying effective chemotherapy drugs⁵⁴. It is also observed that MEA experience a decline in mortality rates from PCa while the reverse is observed for MAA⁵⁵, who are 2.5 times more likely to die from the disease, demonstrate greater cases of recurrences and experience shorter disease-free survival⁵⁶. It then begs the question as to why is there not a greater push towards broadening the cell line bank with PCa from MAA.?

To date, currently 5 cells lines exist from MAA, expected to represent this widely heterogenous disease. Additionally, all of these are representative of African American and none from the Caribbean, an area with even greater reported mortality rates of PCa among MAA. Of the 5 cell lines, 2 were obtained through viral transformation and immortalization while the remainder were spontaneously transformed. RC-77T/E PCa cell line was established through transfection with recombinant retroviral construct, LXSN-HPV16E6E7 containing the E6 and E7 genes of HPV-16 and a neomycin resistance gene⁵⁷ while RC165N was established through telomerase (hTERT) immortalization^{58, 59}. Both RC-77T/E and RC165N/E showed androgen dependence and epithelial morphology cells expressed androgen-regulated prostate-specific homobox gene, NKX 3.1, epithelial cell specific cytokeratn 8, androgen receptor (AR), prostate specific antigen (PSA), and p16. MAA PCa 2a and 2b cell lines were derived from a single site bone metastasis in a castration resistant PCa case in an African American male. These cell lines are tumorigenic, expresses AR+ and PSA+ and are models for advance stage PCa⁵². The other spontaneously transformed PCa cell line, E006AA-Parental (Par), was established as the first cell line of African ethnicity which was from organ confined PCa tumor with a Gleason score of 6⁶⁰. E006AA is an androgen dependent cell line which expresses PSA and Androgen receptors having 26 CAG repeats in exon 1⁶⁰. Furthermore, there is a dispute as to whether E006-AA is actucally prostate which could potentially limit the already limited MAA cell line resourse available to researchers.

Beyond understanding the disease, there is also an importance in identifying the cellular and molecular precursors to PCa. Normal prostate cell lines, BPH1, RWPE-1 and RC165N developed from Benign Prostatic Hyperplasia tissue samples^{58, 61, 62} provide such an opportunity. Through the utilization of these normal prostate cell lines, a greater understanding of the epithelial markers of PCa and the biological alterations leading up to PCa have been understood. Furthermore, these normal prostate cell lines have been instrumental in comparative studies using several drugs and compounds in assessing their carcinogenic/inhibitory roles in PCa^{63, 64}. There is a greater need for more normal cell lines to understand the multistep progression of carcinogenesis, especially if we are to understand the reasons underlying the higher incidence and mortality rates of PCa in MAA^{65, 66}.

Since the emergence of “omics,” this has triggered an advent of target specific treatment options for cancers as opposed to traditional chemotherapy⁶⁷. Through focusing on elucidation of molecular composition of tumors and genetic heterogeneity, several target molecules have been identified^{67, 68}. The drawback however, is that target specific treatments can only be used on a small percentage of patients who match the specific profile⁶⁹. As such there is a need for a panel of PCa cell lines which adequately represent the heterogeneity of the disease⁶⁹. Through the establishment of multiple PCa in vitro systems, a greater characterization and manipulation of PCa at different stages can take place contributing to the current void in understanding the disparities between MAA and other races⁴⁶.

3.3.2. Molecular studies

Improved diagnosis and treatment of PCa can be achieved by a personalized approach especially in light of PCa’s extensive variability in behavior coupled with the multitude of treatment options. Technological advances in genetic analyses continue to pave the way for lead biomarkers that may be exploited towards predicting PCa incidence, outcomes and response to therapy⁷⁰. Technologies such as microarray analyses and next generation sequencing have aided in characterizing genetic mutations in PCa tumor tissue which are subsequently believed to produce individualized bellwethers that can guide clinical decision-making⁷¹.

Current methodologies used for diagnosing PCa are: PSA (prostate specific antigen) which can either be administered as PHI (combines total, free and (−2)proPSA) or 4KScore (consists of total PSA, free PSA, intact PSA and klk2); PCA3/DD3; ConfirmMDx (measures the degree of methylation of GSTP1, APC and RASSF1); PCa biopsies and digital rectal examination. Still, PSA screens dominate the PCa diagnostic regime as others are relatively new with ongoing research and have not been fully streamlined across the globe. The advent of PSA screens occurred in 1970 by Ablin et al.,⁷², and its purification was credited to Stamey et al.,⁷³. As research on PSA was primarily focused on MEA⁷⁴, regardless of the belief held by the US Preventive Services Task Force (USPSTF) that there are no differences amid the groups (MAA and MEA in America) as it pertains to the ratio of risks to benefits, one cannot dismiss the notion of a lack of optimum PSA diagnostic translatability to both groups especially in light of the 2.5 times higher PCa mortality rates among MAA compared to MEA^{75, 76}. The PSA test continues to pose significant barriers since PCa diagnosis is challenged by the heightened variability in PCa clinical course, as the disease can range from indolent and latent to aggressive and metastatic. An ideal tumour biomarker that can be epigenetic, genetic or proteomic in nature is able to differentiate between patients with PCa from those without and also provide information on the stage of the tumour that guides the clinician on a course of treatment and the patient’s prognosis. On the contrary, while PSA can aid in early detection, it or PSA-like proteins are expressed in extra-prostatic human tissues and surrounding fluids that ultimately limits the specificity of this biomarker. Furthermore, the inability of the PSA protein to distinguish indolent from aggressive PCa tumours have consequently led to over-diagnosis and over-treatment of men with the disease which begs the questions; ‘exactly how many men are saved from PSA screening?’ and ‘are the emotional and physical consequences on the millions of men who are over-diagnosed and over-treated, worth it?’

An improved understanding at the molecular level as to the drivers of PCa is necessary to better diagnose the disease and design ideal treatment regimens. This drive has led to over 40 genome wide association studies (GWAS) executed over the last decade which have identified approximately 170 common variants to the disease⁷⁷. At the molecular level, these variants can be categorized as hereditary, sporadic or epigenetic, whereby, the latter two can overlap and some noteworthy ones are listed in Table 2 while others are discussed elsewhere^77–85.

Table 2:

showing various hereditary, sporadic and epigenetic associations with PCa incidence primarily sourced from Shen et al.,¹²⁶ and other sources^{19, 127} can be consulted.

Hereditary	Sporadic	Epigenetic
ELAC	TLR4	NKX3.1
RNASEL	CDKN1B	Myc
MSRI	AR	TMPRSS2-ERG
NBSI	CYP17	PTEN
CHEK2	SRD5A2	Ezhz
HOXB13	Vitamin D receptor	miRNAs
HPC1	CDKN1A	Signalling pathways-Akt/mTOR & MAPK
8q24 & 17q	CDKN1B	Oncogene tyrosin kinase
HNF1B	EGFR	Developmental Signalling
17q21	CAV1	RASSF1
PCA3	MSR
12q24	NKX3.1
2p16	c-myc
10911	PTEN
SPARCL1 on 4q22.1	RbE-CAD
PTPRR on 12q15	AR
10p14	ANX7LZTSI
	Fox P1
	KAI1
	GSTP1
	APC
	ETS P53

The common thread observed in these studies is that more than 90% of these investigations are conducted in MEA. As a result, significant strides have been made in the design of serum and urine diagnostic biomarkers, whereby some are currently clinically applicable as mentioned earlier and these were recently reviewed in Sharma et al.,⁸⁶. Some include i) the noncoding RNA, PCA3 that is believed to be overexpressed in 95% of PCa cases in MEA^{78, 86}, presents in high levels in some men but is unaccompanied by malignancy meanwhile additional studies confirmed an association of this biomarker in Chinese men^{87, 88}; ii) the kallikrein panel (comprises of total PSA, free PSA, intact PSA and human kallikrein-related peptidase 2 (KLK2)) was reported to improve the predictive accuracy of PCa detection over PSA alone with a reduction in the biopsy rate by 362 for every 1000 men with elevated PSA⁸⁹ and these studies were primarily conducted on MEA⁸³; iii) TMPRSS2: ERG fusion & β-microseminoprotein, identification of the fusion gene in urine was reported to enhance PCa detection by more than 90% specificity and 94% predictive value⁹⁰, a research that began by Tomlins⁹¹ and later a GWAS study reported in Germany⁹² and across other ethnicities⁹³, confirmed the importance of this biomarker among MEA, conversely, other studies⁹⁴ did not reveal the ethnicity of the men investigated although given the geographical locale, it is believed they too are primarily of a European ancestry and iv) DNA methylation of genes for example a) Glutathione -S-Transferase-Pi (GSTP1)⁹⁵ that is involved in DNA detoxification and cell cycle regulation, to the best of our knowledge no GWAS research has been conducted on this epigenetic biomarker but smaller studies conducted in Germany⁹⁶, Korea⁹⁷, Italy⁹⁸ and meta-analysis on MAA in America, the Caribbean and in Africa⁹⁹ have confirmed an association of this protein to an elevated risk of Pca; b) the promoter region of adenomatous polyposis coli (APC) that is involved in cell apoptosis, migration and adhesions¹⁰⁰ have been associated with PCa occurrence among Chinese¹⁰¹, Italians¹⁰², Pakistanis¹⁰³ and other ethnicities¹⁰⁴, furthermore, fairly recent meta-analysis¹⁰¹ that covered seven studies (1227 patients for North America and Europe) calculated the pooled hazard ratio to be 1.74 that correlates methylated APC to unfavorable impact on biochemical recurrence of PCa confirming the PCa prognostic potential of this biomarker as well, and; c) the putative tumour suppressor Ras-associated domain family 1A (RASSF1A) whereby meta-analysis conducted on 1123 cases¹⁰⁵ primarily in MEA associated the hypermethylation of CpG islands within the promoter region of these genes to PCa¹⁰⁵. Research has also demonstrated that PCA3¹⁰⁶, TMPRSS2:ERG fusion¹⁰⁷ and PTEN¹⁰⁸ among others⁸⁶ possess PCa prognostic potential although investigations are still quite limited in these areas.

As the awareness increases on the current disparities in PCa incidence and mortality rates, recent times bear witness to up and coming research on the underserved MAA population, albeit limited and on relatively small cohorts primarily from America. Some of these have associated biomarkers, PCA3^{105, 109}, TMPRSS2:ERG¹¹⁰, GSTP1, APC and RASSF1¹¹¹ to PCa presence but more so, the advanced form of the disease. Although some relation of the TMPRSS:ERG fusion gene has been reported to PCa, other research¹¹² has indicated that the fusion gene is twice more prevalent among MEA (49%) compared to MAA (25%) where the latter compared well with Asian men (27%). There are currently no reports on the association of the KLK2 gene to PCa in MAA.

To date, to the best of our knowledge four reported GWAS studies have been conducted with a primary focus on MAA^113–116 from 2007–2016, as 10% of all GWAS PCa studies over the last decade. From these investigations: genetic variant/s in a number of chromosomal regions such as, 8q24 (working in conjunction with another variant) was twice more prevalent in MAA compared to MEA¹¹⁴, a novel risk variant, rs7210100 in the 17q21 region was unique to African-Americans, while its functionality is unknown with respect to PCa, it is reported to be associated to the breast cancer susceptibility gene, BRAC1¹¹⁷; a risk variant in 10p14 was most prevalent in Ghanaian men¹¹⁵ while sub-analysis associated SNPs at 5q31.3 with high Gleason score and finally, Rand et al.,¹¹⁶ reported PCa associations to the gene SPARCL1 on 4q22.1 and PTPRR on 12q15. Other studies reported analyses on varied GWAS research with some focus on MAA^118–121 while auxiliary research have associated additional variants, primarily on 8q24^{8, 122–125}; 10911⁸, 12q24 and 2p16¹²⁶ chromosomes to PCa occurrence in MAA: In identifying these various risk variants be them hereditary, sporadic or epigenetic in nature, it is believed that unearthing their functionality is the key to enhanced PCa diagnosis, prognosis and theranostics among MAA.

3.3.3. In vivo

The use of in vivo models stemmed from the understanding that cell lines in general cannot replicate all disease mechanisms in the original tumours. To understand these mechanisms, specifically for PCa, preclinical models are needed to reproduce tumor growth conditions that mirror the micro and macro-environment of the tumour in the host. Most PCa in vivo systems are modelled using mice due to the favorable genetic manipulations as compared to the other animal models^{127, 128} and extensively reviewed in¹²⁹.

As cell lines are injected in animal models for preclinical assessment, logically, preclinical anti-PCa research would primarily focus on MEA from which most PCa cell lines are derived. The Prostate Cancer Foundation held a Prostate Cancer Models Working Group (PCMWG) Summit in 2007¹³⁰ to address the lack of applicable preclinical models to study PCa tumorigenesis and to advance propitious prevention and therapeutic strategies. The PCMWG identified several limitations in preclinical research including cell lines, furthermore, it was observed that insufficient models with inadequate molecular and biologic diversity were available to reflect human cancers. Indeed, it is reported that only 5% of drugs investigated at the preclinical level using mice are approved by FDA¹³¹. However, it is believed that these statistics could be improved with genetically engineered mouse models (GEMM)¹³². In addition, GEMM should fast track developing individualized therapies for PCa, bearing in mind that the mouse models encapsulate PCa tumours from varied ethnic backgrounds. The African American PCa cell lines E006AA-Parental (Par), and RC-77T/E cell line were both highly tumorigenic in SCID mice^{58, 60} while the former was also tumorigenic in athymic nude mice. The presence of these models maybe utilized in the GEMM paving the way for personalized treatments for PCa among MAA bearing in mind the type of cell line E006AA actually is.

3.3.4. Clinical

Clinical treatment of PCa that depends on age, stage of disease, stratified risk-group and patient’s wishes encompasses one or more of the following: radical prostatectomy (RP), radiotherapy, active surveillance, ADT, chemotherapy among other agents. Patients with very-low or low-risk localized PCa may be managed with active surveillance while locally advanced or metastatic PCa may be treated with androgen deprivation therapy (ADT) or chemotherapy.

Normally, RP is the gold-standard for managing localized PCa and is performed by an open retropubic or perineal approach. Studies evaluating the racial disparities on the oncological outcomes after RP are few in number; many including small numbers of MAA, have short follow-up periods and have conflicting results. The oncological outcomes of RP are dependent on several factors and in MAA, aggressive and more advanced disease at diagnosis are significant¹³³. Of note, National Cancer Registry data suggests that MAA are less likely than MEA to have a RP (22.3% versus 29.1%) and are more likely to opt for no treatment (29.1% versus 23.9%)¹³⁴. This disparity is not clearly understood but may be due to racial or socioeconomic discrimination, MAA mistrust of the healthcare system or more advanced presentation of MAA at diagnosis which precludes curative treatment. Faisal et al reported on 15,993 patients who underwent RP between 1992 and 2013 at a single center¹³⁵. MAA with very-low risk and low-risk PCa were more likely to have adverse pathological features and increase in Gleason score than MEA. MAA with low-risk prostate cancer were more likely to have biochemical recurrence. The Shared Equal Access Regional Centre (SEARCH) Database Study Group analyzed data from 1665 MAA and 2791 MEA who underwent RP in equal-access hospitals¹³⁶. At a median follow-up of 8.5 years, on univariate analysis, MAA had higher rates of biochemical recurrence but reduced overall death. On multivariate analysis, black race was not associated with increased risk of aggressive disease recurrence, metastases or PCa- specific death. An international comparison of African-American (AA), Afro-Caribbean from Jamaica and Caucasian-American men with PCa that had undergone RP reported that AA and Afro-Caribbean men have lower 5-year biochemical-free survival rates than Caucasian-American men¹³⁷. There appears to be no significant differences in quality-of-life outcomes (sexual function, urinary continence) in MAA and MEA undergoing RP¹³⁸.

Radiotherapy may be delivered via external beam therapy or brachytherapy. Both radiotherapy and RP may be used for patients with low-risk PCa and there is a comparable 10-year survival in patients with localized PCa for both¹³⁹. Race does not appear to have a significant outcome in MAA who are treated with external beam radiation^{140, 141}. Results of the Radiation Oncology Group trials reported that factors such as: clinical stage, Gleason score, PSA and lymph node status affect overall and disease-specific survival¹⁴². In this phase 3 randomized trial, MAA had lower overall and disease-specific survival on univariate analysis. However, on multivariate analysis after adjusting for risk categories and duration of adjuvant ADT, race no longer was associated with outcome. There are conflicting results regarding oncological outcomes in MAA with PCa treated with brachytherapy^{143, 144}.

Active surveillance spares men with low-risk and low-volume PCa (clinically insignificant), that will unlikely progress, of treatment-related morbidity. The surveillance protocol varies between centers; however, it is usually onerous requiring frequent PSA testing, DRE and repeat prostate biopsies at intervals. Since increasing numbers of studies have reported that MAA with very-low risk and low risk PCa have pathological upgrading and upstaging after RP, there are concerns about the suitability of active surveillance in these patients (Gocke MI, Prostate Cancer Pros Diseases, 2017). In addition, MAA with low-risk PCa have higher PCa-specific mortality when treated compared to MEA¹⁴⁵. However, MAA continue to be enrolled in active surveillance protocols because there are other studies showing no significant upgrading or upstaging after RP^{146, 147}. Nonetheless, analysis of active surveillance cohorts report that MAA are more likely to have progression and require active treatment^{148, 149}. Further studies are required to evaluate this association since very few studies have been published and these largely have very small numbers of MAA. Interestingly, data from the Surveillance Epidemiology and End-Results Program (SEER)-Medicare database revealed that MAA were less likely than MEA to be placed on active surveillance. Those who were placed on active surveillance more commonly defaulted to watchful waiting after a period of surveillance¹⁵⁰. The reasons for lack of adherence to the protocol could be related to the rigid and prolonged nature of follow-up or even socioeconomic reasons.

Androgen deprivation therapy (ADT) may be used in men with locally advanced or metastatic PCa, as an adjuvant to external beam radiation, for treatment of biochemical recurrence after failure of local treatment or in men with localized PCa who refuse curative therapy. ADT was historically initially achieved through surgical castration (bilateral orchiectomy) or chemical castration by using oestrogens¹⁵¹. Oestrogens are no longer used widespread as they are associated with increased thromboembolic risks¹⁵². Though surgical castration is associated with poorer health quality of life scores than chemical castration, it is still commonly used for management of metastatic PCa. Common first-line methods of ADT include gonadotropin-releasing hormone (GnRH) agonists, GnRH antagonists and anti-androgens. These agents either ultimately suppress testosterone levels into castrate ranges (<50 ng/dL) or inhibit androgen receptors. Biochemical or clinical response to ADT lasts for a duration of 24–36 months, thereafter patients develop castrate-resistant prostate cancer¹⁵³. There are several studies that suggest there are no differences in outcomes of MAA with PCa treated with ADT compared to MEA^{154, 155}. However, analysis of data from the Southwestern Oncology Group phase 2 Study in which both MAA and MEA with metastatic PCa were randomized to orchiectomy with or without flutamide revealed that MAA had poorer overall survival than MEA¹⁵⁶. This statistically significant result was seen even after adjusting for variables such as patient age, Gleason score, PSA and performance status. Contrastingly, Sassani et al¹⁵⁷ evaluated biochemical failure in 681 men from the Southern California Cancer registry (treated in an equal access health care system) treated primarily with ADT, found that MAA had a 34% lower rate of biochemical failure than MEA. The authors theorized that the lower rate of biochemical failure in MAA were due to racial differences in sensitivity of the androgen receptor. The authors however were not able to report on changes in survival.

Chemotherapy is used in men with metastatic castrate-resistant PCa (CRPC) as well as in men with newly diagnosed androgen sensitive prostate cancer. CRPC occurs when there is clinical or biochemical tumor progression despite having castrate levels of testosterone. Median survival in these men is typically less than 2 years. Docetaxel and cabazitaxel, chemotherapeutic agents which lead to apoptosis, have dramatically changed the management and outcome of men with metastatic CRPC by improving survival. In a phase 3 trial comparing 3 weekly docetaxel and prednisone to mitoxantrone and prednisone, there was a significant survival benefit (2.4 months) demonstrated in patients treated with docetaxel¹⁵⁸. Petrylak et al¹⁵⁹ reported a similar survival benefit of 2 months in a similar subset of men treated with docetaxel and estramustine compared to mitoxantrone and prednisone. This trial consisted of 12% and 15% MAA in both treatment arms, however statistical analysis based on race was not done. Prior to 2004, mitoxantrone and prednisone were the standard of care for these patients for relief of symptoms and palliation but with no survival benefit. Cabazitaxel is currently used as a second-line agent in patients with metastatic CRPC who were previously exposed to docetaxel¹⁶⁰. Other than improved survival, chemotherapy also provides a significant decline in PSA, greater tumor response and relief of pain in men with metastatic CRPC. More recently, docetaxel chemotherapy administered for 6 cycles in men with newly diagnosed androgen sensitive PCa prolongs survival for a median 13 months and delays emergence of castrate resistance disease¹⁶¹. As is seen in many other clinical trials, MAA were under-represented contributing to 9% of the study population.

Additional non-chemotherapeutic drugs used in patients with CRPC include: enzalutamide, abiraterone acetate, apalutamide, sipuleucel-T and radium-223. All agents are Food and Drug Administration (FDA)- approved and prolong survival by a median of 3–5 months. Enzalutamide, an androgen receptor signaling inhibitor may be used in men with metastatic or non-metastatic CRPC with or without prior exposure to docetaxel chemotherapy^162–164. Abiraterone acetate, an androgen biosynthesis inhibitor which blocks cytochrome p450 c17 (CYP 17) also increases survival in men with metastatic PCa and may be used in these patients with or without prior exposure to docetaxel chemotherapy¹⁶⁵. Apalutamide is a competitive inhibitor of the androgen receptor and may be used in non-metastatic and metastatic CRPC¹⁶⁶. Sipuleucel-T is a form of immunotherapy for PCa consisting of autologous peripheral blood mononuclear cells. It is administered to men with asymptomatic or minimally symptomatic metastatic CRPC¹⁶⁷. Radium- 223 dichloride emits alpha particles which binds to areas of metastases in bone and is cytotoxic in these areas¹⁶⁸.

Recent advances in drug treatment seek to discover new agents to treat CRPC that are more effective in prolonging survival than current agents and are more tolerable. Although many of these new agents have been investigated, many are not yet FDA- approved. Interestingly, none of these agents have been tested specifically in large populations of MAA. Some include androgen synthesis inhibitors such as the CYP17 inhibitors, VT-464 (Viamet) and TAK-700 (Orteronel). VT-464 produced a greater suppression of the androgen receptor axis than abiraterone acetate in CRPC cell lines¹⁶⁹. The drug is currently being tested in a phase 1/2 trial. Orteronel showed promising results in phase 2 trials with good PSA and testosterone suppression and reduction of circulating tumor cells¹⁷⁰. However, the drug showed no improvement in overall survival in men with chemotherapy-naïve metastatic CRPC and was associated with significant toxicity. The drug is therefore no longer being developed for men with metastatic CRPC¹⁷¹. Androgen receptor antagonists, ODM-201(Darolutamide) which appears to be more tolerable than enzalutamide, does not cross the blood-brain barrier and has low risk of seizures. Phase 1 and 2 trials suggest that the drug is effective in men with progressive metastatic CRPC and is currently being investigated in a phase 3 trial (ARAMIS) including men with non-metastatic CRPC^{172, 173}. AZD3514 is an androgen-dependent and independent receptor inhibitor administered orally. Phase 1 trial showed modest tumor activity but this was associated with significant nausea and vomiting¹⁷⁴. TOK-001 (Galeterone) is a dual androgen synthesis inhibitor (CYP17 inhibitor), androgen receptor antagonist and down-regulator of androgen receptor expression. Phase 1 and 2 trials show galaterone to be effective and safe in men with metastatic and non-metastatic PCa¹⁷⁵.

New promising immunotherapies include pox-based virus vaccines (PROSTVAC-VF), currently being evaluated in a phase 3 trial and Ipilimumab, a human monoclonal antibody administered in combination with radiotherapy^176–178. Additional new agents include Cabozantinib or XL-184, a receptor tyrosine kinase inhibitor of c-Met and vascular endothelial growth factor (VEGF) which appears to be beneficial in bone metastases in PCa¹⁷⁹. Custirsen or OGX-011 is an oligonucleotide which targets clusterin, which is thought to mediate resistance to docetaxel. Most recent data suggests that the agent does not appear to improve survival in patients treated with docetaxel or cabazitaxel^{180, 181}. Tasquinimod, an oral quinolone with anti-angiogenic properties is associated with improved radiological progression free survival in men with metastatic CRPC¹⁸².

4. Overcoming these

The heterogeneity of PCa makes treating the disease a continuous challenge for medical practitioners. Indeed, research findings show an overall decline in death rates from PCa among one group, MEA, more than another group, MAA. Hetergenity of the disease is mutifactoral and takes into account inter alia geographic locale and racial ancestry. Notwithstanding, research shows that the higher incidence and mortality rates among MAA is caused by numerous factors that range from biological to SES. Interestingly, while many instances associate inacessibility to quality healthcare as a contributing factor to elevated mortality rates among MAA with PCa, it is shown that in instances where quality healthcare exist for this group, MAA still have worst outcomes from PCa than MEA. Such findings highlight the signifcant role that biology plays in enhancing one’s understanding of the disease among MAA before treating or curing become possibilities. With this in mind, basic science has to be mobilized at all levels, in vitro and in vivo, whereby access to a wide array of the requisite tools, cell lines, that are used to understand the biological composition of the myriad of PCa tumours from MAA, can be used. These tools in addition to tissue samples from the original tumours can be exploited for WGS, paving the way for identifying biomarkers that can be used to improve PCa diagnoses among MAA while developing theranostics for this group with the disease. What is clear is that these studies have been limited to MEA and this is believed to contribute to the reduction in PCa mortality rates among this group. Moreover, clinical trials are primarily conducted around MEA further concretizing the reason for reduced PCa mortality rates among them. Certainly, tackling the burden of PCa among MAA has to engage relevant stakeholders from public to private sectors to scientists. But importantly to note is that while earlier diagnoses can inadvertently curtail PCa progression among MAA and thus optimize treatment, the advanced stages of the disease are more often associated with the increased mortality rates among them. Hence, more tailored chemotherapies are required that MAA can equally benefit from compared to MEA and this can only be achieved through research at clinical and preclinical levels. An intentioanl shift in the way we conduct research has to incorporate equal representation among all groups with PCa. Social barriers have to be broken down rendering equal access to quality care for all groups.