Racial Disparities in Prostate Cancer Outcomes: Biology or Society?

The tumults of 2020 have put racial disparities front and center for the nation at large and health disparities are now viewed with increasing intensity, if not clarity. Prostate cancer patients and the treating community have struggled with the question of racial disparities in outcome for some time, and recent events bring this closer to the front of our dialog.

As part of a busy academic year, I have been part of or launched several research projects, written or helped write several large grants, and even served on a few grant review panels. In doing so I have been involved in many conversations around making research questions have a disparities ‘angle’. This has been suggested to researchers for a variety of reasons – doing so will increase the likelihood of funding, doing so will identify drivers of biological differences in the disease, and doing so will identify gaps in our care that we can work on closing. All are likely correct, even the slightly cynical suggestion that adding disparities questions to grants helps increase the likelihood of fund. I can verify that these questions DO come on grant review panels, and are viewed as important.

But, when mentioned, I am struck by the diverse perspectives that reviewers, researchers, patients, and patient advocates take when the question of disparities arise. Therefore, in order to improve my own clarity of thought (honestly, that is frequently the reason I start these blog posts!) I have endeavored to ‘frame’ the questions of disparities in a few discrete, and hopefully addressable, categories.

As I see it the questions are:

1. Differences in outcome
   1. Are differences in outcome evenly distributed by the stage of disease?
2. Differences by host
   1. Do host factors drive significant differences, by race, in the outcome of prostate cancer?
3. Difference in disease?
   1. Are there differences in actionable genomic alterations in the tumors?
4. What do we know about our experimental tumor models and their relationship to race?
5. Are differences in response to treatment observed? If so, are these differences due to pharmacokinetic, financial, or other factors, such as adherence to therapy or systems of care?

In a recent letter to the New England Journal of Medicine, Brandon Mahal from the University of Miami, and collaborators from the Dana Farber Cancer Institute and Memorial Sloan Kettering Cancer Center described an analysis of the differential genomic findings between Caucasian, Black, and Asian prostate cancer patients.  They analyzed approximately 474 cancer-associated genes with known mutations across 2400 patient samples that had undergone next-generation sequencing at these institutions. Of the samples they analyzed, about 1400 were from primary tumors and the rest from metastases. Among the metastatic tumors, 801/70/37 percent were from White/Black ( they used Black as a descriptor, not African American) and Asian patients, respectively. Results were a bit surprising and here are a few of the key takeaways:

1. Black patients are more likely to harbor DNA repair mutations of any kind.
2. Black patients with metastatic disease were more likely to have actionable mutations.
3. Androgen receptor mutations are rare but more prevalent in Black patients than the others.
4. Seven percent of Black patients had BRAF mutations, which are prevalent in other cancers and are targetable.
5. TP53 mutations are much more common in Asian patients than in the other groups.
6. Asian patients are far more likely to harbor FOXA1 mutations in the primary tumor

There are a few problems with these data: confidence intervals were not adjusted for multiple comparisons so we should only regard differences with very low P values as significant, there is no readout of BRCA2, perhaps the most common actionable mutation that is studied, and the samples were only derived from a handful of centers. Nevertheless, they do point towards a few key areas for further study – notably the relationship between TP53 alterations and metastasis in Asian men, the rare but potentially targetable finding of BRAF alterations in Black men. – detected in 7% versus only 1.5% in White men. BRAF mutations are commonly found in malignant melanoma, lung cancer, and colorectal cancer and are targeted with the drugs vemurafenib and dabrafenib. If confirmed (and if these are indeed pathogenic mutations) a new path to treatment selection may have been discovered in this analysis.

Similarly, Franklin Huang, now at UCSF, has explored the potential for key biological driver differences that may distinguish prostate cancer in African Americans from White patients.  After reviewing several hundred samples from African Americans and European prostate cancer patients, they conclude that frequencies of common genomic alterations were not significantly different between the groups  - although they did make a few key observations:

1. MYC amplification is more common in African American Men than Europeans
2. PTEN loss was less common in African American men than Europeans.¹

Overall, however, rather than state that this represents a failure to find a difference, they suggest that this is an opportunity for understanding that our standard precision medicine tests are likely equally beneficial for patients, regardless of race. Mutation analysis is an important step and represents something that all of us in the clinical care of medicine can, and should, be doing on a regular basis in all of our patients.

In short, there is much more that unites us than what separates us.

Beyond DNA, one wonders if there are distinct pathways of activation in African American patients that may drive disparities.  Saleem Bhat and his team from the Department of Urology here at the University of Minnesota have cultivated cell lines from Black patients and compared them to those derived from White patients. In a recent paper, they described the outcome of men with loss of ROBO1, a transmembrane protein that binds and suppresses DOCK, a key driver of cytoskeleton motility and thus, metastatic spread. Their work suggests that men with ROBO loss in primary tumors are more likely to go on to develop metastatic disease after surgery. Further, in experimental model cell lines derived from African American patients and comparing them two cell lines from white patients, they show a key difference. Although ROBO1 levels in tumors from African American versus White patients were similar, epigenetic processes may reduce ROBO1 expression and function: The African American cells were more likely to be silenced through methylation that the tumors from White patients. Further, they showed that ROBO1 loss leads to greater cell motility – a key and necessary process for the development of metastasis.^2,3 These are early data that require further validation.

IF biological differences are only modest, why have we observed for so long that clinical outcomes are vastly different?  My read on these data is that because the biological differences are present but modest, societal concerns may override them.  In support of this point are a series of very clear clinical outcome papers over the past few years, and to me the most striking is that by Dess and Colleagues in JAMA Oncology last year.^4,5 In this analysis, the team looked at differences in outcome in three distinct cohorts: SEER, which is US Medicare patients and based on claims across the whole country without regard to finance, geography, or treatment; the Veterans Administration – wherein patients all receive care within the same nationalized system; or on randomized controlled trials, where the treatment is very tightly regulated down to dose, stage and outcome analysis. What the authors report is that African American patients experience a slightly worse outcome (0.5% greater prostate cancer-specific mortality at 10 years) in the SEER cohort, no difference in outcome when treated in the VA system, and an improved outcome ( compared to White men) in the randomized controlled trials.

These and other data put us on notice. Biological differences exist but may be washed away when we address treatment disparities. All of this underscores the need for broad inclusion in epidemiological studies, clinical trials, retrospective analyses, and yes, grant proposals.

References:
1. Koga Y, Song H, Chalmers ZR, et al. Genomic Profiling of Prostate Cancers from Men with African and European Ancestry. Clin Cancer Res. 2020 Sep 1;26(17):4651-4660.
2. Ferrari MG, Ganaie AA, Shabenah A, et al. Identifying and treating ROBO1(-ve) /DOCK1(+ve) prostate cancer: An aggressive cancer subtype prevalent in African American patients. Prostate. 2020 Sep;80(13):1045-1057.
3. Parray A, Siddique HR, Kuriger JK, et al. ROBO1, a tumor suppressor and critical molecular barrier for localized tumor cells to acquire invasive phenotype: study in African-American and Caucasian prostate cancer models. Int J Cancer. 2014 Dec 1;135(11):2493-506.
4. Dess RT, Mahal BA, Spratt DE. Biology vs Access to Care-Relative Contribution to Racial Disparities in Prostate Cancer-In Reply. JAMA Oncol. 2019 Oct 31.
5. Mahal BA, Alshalalfa M, Kensler KH, et al. Racial Differences in Genomic Profiling of Prostate Cancer. N Engl J Med. 2020 Sep 10;383(11):1083-1085.