Genomic Comparison Between African American Men & European American Men with Prostate Cancer - Isra Elhussin

December 19, 2022

Charles Ryan and Isra Elhussin engage in an enlightening conversation about the disparities in cancer genomic profiling between African and European ancestries. Dr. Elhussin's pioneering research investigates how genetic ancestry and specific gene mutations, notably SPOP mutations more prevalent in those of African ancestry, influence prostate cancer subtypes and treatment responses. Her study highlights higher PD-L1+ and CD8 population co-expression in the African ancestry, suggesting potential for targeted therapies. The discussion further underscores the necessity for inclusivity in clinical trials to accurately represent the African population and better interpret the clinical relevance of these findings. The conversation underlines the complexities of cancer incidence and mortality rates among different populations, hinting at the need to move beyond race and focus on molecular subtypes in cancer studies, thereby paving the way for more personalized and effective treatment strategies.


Isra A. Elhussin, MD, MSc, PhD, Cancer Bioinformatician, Tuskegee University, Auburn, Alabama

Charles J. Ryan, MD, the Chief Executive Officer of The Prostate Cancer Foundation (PCF), the world’s leading philanthropic organization dedicated to funding life-saving prostate cancer research. Charles J. Ryan is an internationally recognized genitourinary (GU) oncologist with expertise in the biology and treatment of advanced prostate cancer. Dr. Ryan joined the PCF from the University of Minnesota, Minneapolis, where he served as Director of the Hematology, Oncology, and Transplantation Division in the Department of Medicine. He also served as Associate Director for Clinical Research in the Masonic Cancer Center and held the B.J. Kennedy Chair in Clinical Medical Oncology.

Read the Full Video Transcript

Charles Ryan: Hello, I'm delighted today to speak with Isra Elhussin, MBDS. Dr Elhussin is actually a PhD candidate at Tuskegee University in Tuskegee, Alabama and will soon be doing a postdoctoral at Johns Hopkins. She was recently named as a next gen star of the American Association of Cancer Research and I think we're going to see why as we talk to her about her work. She is a bioinformatician and is going to talk today about cancer genomic profiling with a particular comparison between African ancestry and European ancestry. And then we'll have a conversation on the other end. Dr Elhussin, very nice to see you.

Isra Elhussin: Thank you Dr. Ryan for the nice introduction. Thank you for having me today and I'm glad to be here and share some of our findings. So as Dr. Ryan mentioned, my title today will be Cancer Genomic Profiling Comparison between African Ancestry and European Ancestry.

So at the beginning, I want to start with highlighting some of the cancer disparities where we see higher incidents and mortality rate among black population. And with the prostate and breast cancer are on the top of the list with more than 15% overall mortality in African American population.

So now when we take a global view, prostate cancer incidents in North America is one of the highest worldwide. And when we are looking back to the trans-Atlantic slave trade, people were transported as a slave from West Africa to North America and globally. And this suggests some of the possible ancestral basis and some of the shared outcome.

So there's multiple factors that contribute to cancer disparity and such as genetic ancestry, such as psychological, such as stress, depression, as well as access to healthcare and socioeconomic status, which has directly into the survival disparity that we see within the African population. Also, there is the biological process such as immune response and therapeutic resistance, as well as the behavioral process.

And currently, we added tumor promoting inflammation, as well as the genomic stability and mutation to the cancer hallmark. So racial disparity in prostate cancer, also linked directly to the aggressive subtype, which will give you adverse outcome.

From the TCGA was defined the molecular taxonomy of prostate cancer and they have divide the prostate cancer to seven subtype. One of them is the ERG status as well as SPOP, FOXA1 and the ETS family. And where we see that ERG status are mutually exclusive with SPOP mutation. Van Allen group also confirmed that SPOP status and ERG status are mutually exclusive in a larger population.

So before we move forward, we need to clarify, what is the difference between race and ancestry. So race is built on non-biologic characteristics such as skin color, and that we have the old concept of five race, while ancestry have more highlighting the actual genetic variation, your parents and grandparents, as well as the geographical location for the population.

Really now, we need to define, who am I? And that's why we tried taking the ancestry approach and we did bioinformatic analysis using global ancestry and local ancestry. And the global ancestry have super and subpopulation. The super population have five super population based on the thousand genome database, which is, this is the different database that we use currently in the bioinformatic analysis. We have African, we have European, we have South Asian, East Asian, and ad mixed American. And each one of these super population have their own subpopulation. But today I'm going to focus on the African subpopulation.

The African subpopulation have different subpopulation through Africa and also we have the African Caribbean ancestry and American of African ancestry in Southwest USA. These two, the subpopulation didn't define what is the African descent, specifically in Africa. While the others we have, they were highlighting, we have from Nigeria, Europa and Isan and we have from East Africa, we have Kenya. And we all give you the descent of those populations, specifically in Africa. While the local ancestry will highlight the specific ancestry origin of a specific chromosome segmentation where you have more ad mixed population.

So we have our total for this specific study that we're going to present today, we have total of 72 patients, 16 where self-reported it as African American, 25 we're self brought it as European American, where we have 30 the patient with unreported race. And really here you can see the importance of the genetic ancestry, because even now it will highlight the biological and the actual genetic variation. As well as if you have unreported self-reported race, then you can cover or estimate the ancestry. So when we did the ancestry estimation, we re-characterize our patient as 28 as African ancestry, 37 as European ancestry, 5 were other such as East Asian, South Asian, and ad mixed American and 2 were ad mixed ancestry. So we did DNA and genomic study, whole exome sequencing, as well as RNA sequencing. We have totals 54 patient as whole exome sequence and 65 patient as the transcriptomic analysis, and 47 patient will have match both. So we have as a multi-omic layer in here as 47 patient.

So most of our patient who self-reported their risk as African American, they have larger proportion of the African ancestry. And now those specific patient they descended from, most of them as you can see here from ASW which is African in Southwest USA and African Caribbean.

So really the 1000 Genome database didn't answer our question, where those patient really descended in Africa? So that is highlighting the underrepresentation of African population within the reference database, that currently now highlighting more European patient. So to untackle that we use another database called HGDP which is has more representation for the African genome. And when we run the subpopulation analysis we found those patient dissented specifically from Bantu population or Europe population.

So now the question is, what the ancestry associated transcriptome signature specific for those patient? We found that when we run, analyze the RNA sequence data, found that there is 300 gene, were enriched in the African ancestry, differentially expressed in African ancestry patient. And gene enrichment analysis for those patients were male immune signaling pathway. And when we look at the interferon signaling pathway, which is one of those, the pathways that we are enriched in the African ancestry, we found that these immune signaling genes were highly correlated with the African ancestry.

And now we want to know, what specific immune cell that enriched in those transcriptomic data. And we use CIBERSORTx to measure the abundance of the immune cell from the gene expression data. And we found that the immune score, immune cell score, were significantly high in African ancestry compared to European ancestry and specifically were T-cell.

So now we want to look if there is specific genomic alteration or genomic deliver for this transcriptomic signature. And when we run the analyze the whole exome data, we found that SPOP mutation and FOXA mutation were on the 30 top mutations. Which is, as I mentioned earlier, those are two of the subtype of prostate cancer that were reported by TCGA data. So now we look if that SPOP mutation is associated with African ancestry or not. In our data, you can see the trend, but we have small sample size, that's why we couldn't see the significance. But when we confirm it in the TCGA data, we found that SPOP mutation is significantly associated with the African ancestry population.

So from there, we looked what is the percentage, we divided the whole exome data now for African ancestry and found that SPOP is 20% in our African ancestry population, versus 10% in the European ancestry.

And now we want to look specifically for what residue or variant that within that SPOP mutation. So there is several residue that reported from other groups in the field, and when we look in our patient we found that there is different residue and variant between African ancestry and European ancestry. And since we have small size data that right now we are collecting more sample size and we have our clinician from John Hopkins that's helping us with increasing our sample size. But meanwhile, we were thinking, does those variants are specific to the African ancestry or not?

So we run the local ancestry and we choose two patients with dominant global ancestry and dominant global European ancestry, and we found that chromosome 17, this is where the genetic location for the SPOP mutation that is linked if you have a dominant global ancestry, either European or African, you will have 100% at that segment for SPOP. So that means that those specific variant, that's where unique to African or European ancestry, also confirm with the local ancestry at that segment that is 100% for each African or European ancestry.

Now we brought another patient who has SPOP mutation but has ad mixed global ancestry and we looked at chromosome 17 again, and we found at that specific location that 100% that patient have African ancestry. So it does mean that the SPOP mutation could be specific link to African ancestry or unique to African ancestry. And that is really promising because currently there's multiple group who reported that prostate cancer with SPOP mutation patient, they have better response to androgen targeted therapy compared to the standard treatment of therapy which is docetaxel.

So again, I mentioned during the literature review that SOP mutation is mutually exclusive with ERG status. So we run the fusion status for our patient and we found that all the patient with SPOP mutation are within the ERG negative status. And now we run those transcriptomic signature for based on the ERG status, SPOP mutation, as well as ancestry. And you can see that the spectrum of the transcriptome signature is changed with the African ancestry percentage, or so if you have high African ancestry you will have the signature more high. And we were asking what specific type of signature in there within those patients.

So when we run the gen enrichment analysis we found that the interferon signaling pathway as well as other immune signaling pathways, such as PD1 signaling pathway, were enriched in the African ancestry with SPOP mutation compared to European ancestry with SPOP mutation. But now we want also to look if those African ancestor with wild type have similar compared to the RPN with wild type similar signature. And we found that there is some of the immune signature in there but is different immune signaling pathway.

So these are some of the SPOP service rate that we collected from the literature and reported from before that they are direct substrate to SPOP. And we found that most of those substrate are highly correlated with the African ancestry and we were interested in PD-L1, because PD-L1 one as you know have directly to the interferon signaling pathway which were enriched in the SPOP mutant African ancestry patient.

So we start with asking the question, are really African tumor are highly immunogenic or not? And then we characterize our transcriptomic immunogenic data based on the immunogenic which is high PD-L1 and then non immunogenic, which is low PD-L1. And we found that nearly that African ancestry have higher expression of PD-L1. And then within that African ancestry we look for the expression of the CD8 and PD1, I mean within the PD-L1 population or the immunogenic population. And you can see there's some trend in there that CD8 and PD1 are co-expressed with within PD-L1 or immuno type are more in the African ancestry patient.

So we went to the TCGA data and now we look if there is any correlation between PD-L1 and immune cell within the TCGA data and there is different correlation between PD-L1 and other immune cell. But we were interested specifically in CD8 T cell.

We did special transcriptomic analysis technology, we use Akoya bioscience to confirm if there is any really African ancestry tumor are hot or not. Because as you know all that in the scientific community right now, that the prostate cancers is called tumor. But most of these the study were done in European ancestry patient. So now the question is, is the African ancestry behave the same tumor, behave the same as the European ancestry or not? So we did the special transcriptomic and as you can see here, and this was low grade tumor of African American and European American, as you can see here, there is infiltration of the CD8 positive PD1 positive within PD-L1 population, in compared to the European American tumor where you can see most of it is in the tumor, in the stroma not infiltrated to the tumor.

And we did segmentation based on PD-L1 population and we found that and we comparing between both patient and we found that CD8 population co the breast with PD1 population within PD-L1 population. And as you can see here is more higher in the African American compared to the European American patient. So then we did a quantification for our protein and we found that PD-L1 population are very significantly high in African ancestry compared to European ancestry. And when we look at the CD8 positive PD1 positive within that PD-L1 population we found that is higher in African ancestry compared to European ancestry.

So in summary, 70 to 85% of our African ancestry are descended from Bantu and Yoruba from West Africa. And 15 to 30% were European ancestry descended from Orcadians in Scotland or Tuscan in Italy. And that's African ancestry associated with ERG mutations, SPOP mutation with high immune inflammation signature and decrease overall immune response due to T-cell exhaustion. And we link that SPOP mutation linked to the highly immunogenic subtype as well as CD8, PD1 immune infiltration.

So in summary we have that SPOP mutation are mutually exclusive with ERG negative as well as, from the literature reported that high associated with high spin one as well as high PI3K, mTOR and indirectly related to CHD1 deletion. Also SPOP mutation or loss of function will lead to accumulation of the subrate direct substrate such as AR, SRC3, PD-L1. And we are interested in specifically in PD-L1 where you see that is direct substrate of SPOP and that will lead to aggressive subtype of prostate cancer in African ancestry. SPOP also linked to high interferon signaling pathway and high as well as higher in African ancestry. And so SPOP mutation could be a promising target and biomarker for prostate cancer since as I mentioned earlier that right now there is a lot of group that clinicians study SPOP mutation could be as immune target through the PD-L1 or immune checkpoint and AR targeted therapy.

At the end I want to thank my lab and my mentor Clayton Yates and all the Yate's lab, as well as our collaborations Stefan Ambs from NIH/NCI. Isaac Kim from Yale University, Moray Campbell from Ohio State, Melissa Davis lab as well from Weill Cornell. Emmanuel Antonarakis from University of Minnesota, Tamar Lotan and Cathy Marshall from John Hopkins. And for our funding DoD as well as NIH and Prostate Cancer Foundation. Thank you.

Charles Ryan: Thank you. Great presentation and really interesting data. Let's tease this out a little bit for those who are hearing it for the first time. So essentially you're showing that SPOP mutations are more common in an African descent population. So that's point number one. There's an emerging body of literature to show that SPOP mutations have a better prognosis overall and probably a better response to androgen deprivation therapy. But at the same time what you're suggesting is that SPOP mutations drive a high PD-L1 expression which actually may be sort of paradoxical in that it may lead to resistance to immunotherapy. Am I reading that correct?

Isra Elhussin: I'm saying that will be if you have high PD-L1, that you could be a candidate for immunotherapy or immune checkpoint.

Charles Ryan: Right.

Isra Elhussin: And that's really will... We need more African population in the clinical trial to test that. Because currently most of these immune checkpoint and immune therapy are tested in European population. So I think the better representation of the African population in clinical trial that's will give us the answer.

Charles Ryan: Certainly, we would all agree that that would be a great thing for us to do in the field. PD-L1 expression is a little complicated, because in certain cancers it has not really been shown to be associated with benefit for immunotherapy. And we don't use the check one inhibitors in prostate cancer in a widespread manner. So it's a little complicated of how we would be able to move forward with that. But I think the SPOP mutation story is actually quite fascinating from the standpoint of what we might be able to do for treatment deintensification for certain individuals.

So I'm going to highlight a couple of pieces of literature you're probably aware of. First of all, getting back to the immune piece, is with CPU lus T and immunotherapy that is used in certain settings in prostate cancer, there have been studies that have shown fairly consistently that African American men who get CPU lus T have a better outcome than European descent individuals who receive that same therapy. Do you think your findings could support that as those findings as well?

Isra Elhussin: Yeah, I think so.

As you mentioned also, it's different according to the patient, according to their genetic app makeup, according to their tumor microenvironment. So we really will never know until we try this in a larger population. And currently, I think maybe you are aware about the case from Australia, I think she just published global cancer inclusion and include a lot of sub-Saharan African ancestry population. And they were looking specifically for highlighting new taxonomy for prostate cancer. And from there, they have the global mutational subtypes, and there is two subtype that are linked directly to African ancestry. And really this largest study, and they have around 185 patient from African ancestry.

So this type of a study or genomic study as in a global inclusion could also change our understanding for prostate cancer and subtype and taxonomy. And really what's fascinating in that study, they combine both the subtype or molecular profiling with the genetic ancestry, as well as the epigenetic modifiable factor. So all of that were in the same model at the same time and then they analyze the data. And they have both whole expome, whole genome sequencing and other transcriptomic data. So really this type of study can give us where we go, to lead us where we go and how we go, and what direction I think.

Charles Ryan: So I fully agree that what we need to do is increase the inclusion of African American or individuals of African descending clinical trials. But your second point is also a wonderful one, which is that it shouldn't be so much about race, it should be about molecular subtype. The other piece of literature that's quite interesting, is there have been studies looking at African American outcome, outcome of African American men on various hormonal agents, hormonal therapies in prostate cancer, suggesting that as a population that there is an improvement in an outcome for African Americans compared to the European descent individuals.

Now it's probably being carried by the double, the twofold, increased risk or twofold increased likelihood, I should say, of an SPOP mutation which confers improved outcome with hormonal therapy. So it's really a fascinating issue, which is to say, we say repeatedly, and it's true, on the one hand African American men are more likely to have prostate cancer, they're more likely to develop it and to die of it. But on the other hand, we've found this window for 20% of them where they may actually have an improvement in their outcome if they're given the appropriate therapy. So your work has huge translational potential and adds to that piece of lit, that body of literature very, very nicely. So I want to thank you for your time and thank you. And congratulations again on your award from the AACR and we look forward to hearing further great things as a result of your work.

Isra Elhussin: Thank you. Thank you. Really appreciate this chance and opportunity to meet you, first, and then to have our audience from Euro today. And I really appreciate that, your time and everyone who's working today to make this happen.

Charles Ryan: Very good. Thank you again.

Isra Elhussin: Thank you.