Well, we’ve known for a long time that there are higher rates of genomic alterations in the androgen receptors in the African American community, and not only will that have a potential effect on treatment response, but also in the presentation. There was a very nice abstract in the American Society of Clinical Oncology Genitourinary Cancers Symposium (ASCO-GU) presentation using the Surveillance, Epidemiology, and End Results database to examine racial disparities. As prologue, we’ve been concerned about the US Preventative Services Task Force recommendations against prostate-specific antigen screening. There have been several studies that are not definitive, but they do raise the possibility that patients may be presenting with higher Gleason scores and perhaps metastatic disease. But in this particular study at ASCO, the investigators looked at racial disparities for men in the African American community, and what they found was that these men were significantly worse off – they were presenting with higher rates and with metastatic disease as well. There really is a higher risk of prostate cancer-specific mortality in that population if we don’t screen. I think these are very provocative data that we have to utilize in our recommendations to patients, in terms of screening. I think that goes to the heart of the matter of the question of differences in specific cancer genes that could impact not only treatment response, but if we step back even earlier than that, the screening process.

I hate to cite some of our own work, but I will on this. Familial or hereditary prostate cancer in African Americans represents an area that has been discussed in the literature, but there have not been a lot of studies. There was a PhD in genetics (whom I have actually now hired as a postdoc) who was looking at hereditary prostate cancer in African Americans, and one of the things that she found was really quite intriguing: a linkage to MSH6, which functions in the mismatch repair pathway, a subset of the DNA repair pathway. It’s a little bit intriguing because in the mismatch repair (at least in other cancers; we really haven’t done a lot of studies in prostate), we end up with a much higher mutation rate. That might imply that there’s going to be a higher response to immunotherapy. When it comes to the programmed death-1 (PD-1)/PD ligand-1 (PD-L1) pathway, we really don’t have data on that front. For sipuleucel-T, there’s been a recent sort of re-analysis and aggregation of the data from the sipuleucel-T where it looks like the African American group seemed to respond a little better than expected. In the world of prostate cancer, we’ve kind of gotten accustomed to the idea that African Americans will present with younger age and they’re going to have higher rates of metastatic disease and higher death rates. It has been a whole cacophony of bad news for African Americans. This might be a little piece of good news: it could be that, at least for sipuleucel-T, a patient might respond better if he is an African American. We have other data that we’re actually preparing right now, and it will be presented orally at the American Urological Association. It takes an expanded data set and sort of confirms these initial findings. The idea that maybe African Americans can respond better to immunotherapy is a provocative one, and clearly, we need to get more data.

We did a study at Mount Sinai where we looked at abiraterone specifically in African American patients. The reason we performed this study was that prior studies, including one we had done when I was at Dana-Farber Cancer Institute in Boston, MA, had looked at germline polymorphisms in androgen metabolism genes and found that success with primary androgen-deprivation therapy varied, not just by clinical factors but also by germline polymorphism in metabolism genes for androgen and estrogen. This makes sense because patients basically are born with innate abilities to metabolize androgens, and that may impact the duration of benefit from androgen-targeted treatment. But the problem was, and we published this, that it was very, very hard to recruit patients to our trial even though we are in a population where there are actually quite a number of African American patients in our practice. This leads to a second issue: even though the response rate seemed on the surface to be equivalent to the Caucasian population that comprised the majority of patients in the original abiraterone trials, it’s really hard to recruit patients in general to clinical trials, particularly African American patients. These are really important questions about whether there are, for example, racial or ethnic differences in the efficacy of specific drugs in advanced prostate cancer and other cancers. Yet we often don’t have enough patients to really identify this based on the level 1 studies that lead to the approval of the drugs, or small phase 2 trials that might be performed subsequently. I know that I continue to struggle with understanding these differences because, unfortunately, there are not a lot of data to really help drive our recommendations for subpopulations.

I completely agree that the reality is that this very well may be a different subset of patients. We might need to think about patients differently and how we treat them by race. The more evidence that builds regarding the enrichment or benefit to different types of therapies, whether it be immunotherapies or hormonal therapies, the more we’re going to need to recalibrate when we introduce those agents into the treatment sequence for these patients. I think that’s critical, and then to your point and to Dr Sartor’s point about looking at the mechanisms underlying that, whether they’re germline polymorphisms or mutations, what are the underlying characteristics within this population? Not all African Americans are the same, and there are going to be subsets within there. I like the germline analyses because that’s something that’s available to everybody. It’s really hard to do the tumor biopsies especially early on in metastatic settings, but the germline analysis is something that’s relatively affordable and easily accessible to do in the entire population. That’s another area where we really need to dedicate some efforts. You’re absolutely right: this is a really difficult patient population to recruit to trials for a number of reasons. I think this will remain an area that is going to stay underdeveloped despite a lot of efforts. We may never get level 1 evidence, but we may need, as Dr Sartor said, repeated studies showing the same patterns, even if they are limited, retrospective, or what have you. These are going to be the data that may drive some of changes in the future.