Disparities in Prostate Cancer Management in the United States Presentation - Alicia Morgans
September 4, 2019
At the 2019 Advanced Prostate Cancer Consensus Conference (APCCC) Dr. Alicia Morgans presented on disparities in prostate cancer management in the United States. Dr. Morgans highlights the 174,650 new cases of prostate cancer in the United States in 2019 and noting that social health is determined by many factors which could be responsible for these disparities, she frames her discussion around which Black Americans have approximately 2 times higher incidence of prostate cancer vs. white Americans, and ~2.5 times higher mortality.
Biographies:
Alicia Morgans, MD, MPH Associate Professor of Medicine in the Division of Hematology/Oncology at the Northwestern University Feinberg School of Medicine in Chicago, Illinois.
Biographies:
Alicia Morgans, MD, MPH Associate Professor of Medicine in the Division of Hematology/Oncology at the Northwestern University Feinberg School of Medicine in Chicago, Illinois.
Related Content:
Written Conference Coverage: Disparities in Prostate Cancer Management in the United States
Written Conference Coverage: Disparities in Prostate Cancer Management in the United States
Read the Full Video Transcript
Alicia Morgans: Thank you. And I'm hoping not to hear the cowbell, but I have two small children so I've heard worse. Also, thank you to the organizers for giving me the opportunity to present on the United States and on our disparities. Clearly we can see that around the world there are huge disparities that we all must tackle. This is a small portion of them and I appreciate the time that we're allotted.
This is an outline and I want to say, and we'll discuss that the issues of disparities in the United States are actually bigger than the black and white American disparities that we talk about traditionally and regularly. We'll explore that a little bit. We'll talk about how personalized medicine and the clinical trials that we do can be affected by these disparities and that the data that we have may not be applicable to everyone that we treat. We'll talk about how the access to treatments may be one of the barriers that is definitively impacting these outcomes. And if we improve access, perhaps we can do better. Some ongoing efforts that are trying to make a difference and some conclusions.
I think the most talked about disparity in the United States as you can see here on the incidence curve on the left with the red line showing the incidence in African American men versus other races below and the death curve on the right, demonstrate the black Americans have a two time higher incidence of prostate cancer versus white Americans and two and a half times higher mortality.
And when we think about the causes, I think it's important for us to recognize that it's probably more than just biology. And as I said, I'll show some data to support that, but biology is probably playing some role as well. When we see that the incidence is higher, we don't expect that black men are being screened with PSAs or DREs more commonly than white men. There must be something biologic as well. But the social determinants of health that we see on the right, things like someone's culture, their access to care, financial hardship, their literacy, and numeracy, are actually probably playing a large role in their outcomes as well.
As I mentioned, this is a much bigger story than just black and white. We're really simply scratching the surface when we speak only about these racial issues. There are also disparities by socioeconomic status, age, geography, education and many others and it's also important for us to recognize that disparities exist in therapeutics and in outcomes in terms of disease outcomes but also in supportive care for these patients. And I'll mention that with an example in a few minutes.
On the right is an example of disparities by age that our group put together a few years ago using some SEER data and we saw that as each year went forward, from the 1990s through 2009, there were improvements in terms of prostate cancer-specific survival with each year, but for those patients who were over the age of 75 there was actually no improvement after 1999. This is an example of improvements in survival that are more pronounced for older men. The reasons of which are unclear, but a clear disparity by age.
Another disparity that I'd like to draw our attention to is disparities by education and poverty that can also certainly in the United States, drive a disparity where we have issues of patients not having health insurance and access to healthcare. The Appalachian Mountains or Appalachia is a region that you can see that I've circled here on the right side of the United States. And on the right figure, you can see that these counties are divided up by poverty with red areas demonstrating very desperate poverty and blue areas being areas that have really no economic hardship. In the state of Kentucky, which is circled in that right figure, there's a large area of the Appalachias that are completely economically devastated. As you can see in red.
A group in Kentucky looked at prostate cancer-specific outcomes in men and compared their survival in these men who were from the Appalachian part of Kentucky to those men who were from non-Appalachian Kentucky and found, as you can see in the survival curve on the right, poor survival in men who were from the Appalachian part of the same state. This was associated with lower education attainment, higher poverty and greater number of co-morbid illnesses. Within the same state these issues, and this is a state of predominantly white Americans, these issues, we're driving a very clear disparity.
In California, similar studies were also done. This was a fascinating study done by a group that looked at multiple tumor types and specifically, this is the prostate cancer data. They looked at 270,000 men with prostate cancer and did a mediation analysis to identify which factors were driving disparities by race. If you look at the image on the right, you can see the black Americans were highlighted in blue, white in gray as the reference and black men had poorer survival than other ethnic groups. This was predominantly driven by stage at diagnosis. That was 24% of the disparity indicating again, that I don't think incidence that's higher in black Americans, is actually related to earlier screening. But this was also associated in this analysis with marital status, with neighborhood socioeconomic status and with hospital insurance composition, which really defines whether hospitals are receiving predominantly government supported care or private insurance. And higher rates of things like Medicaid or government supported care were associated with poorer outcomes and we're driving a large part of the disparity in this population in California.
As I mentioned, supportive care also is different or applied differently across patient populations. This is a study that I did with Matthew Smith many years ago when I was one of his fellows and we looked in the SEER Medicare database and found that DEXA screening for bone health, for fragility fractures, was actually quite uncommon for all patients, but was particularly uncommon in patients who were older, over 85, for black Americans and for those patients who had lower educational attainment.
As we know and as we've discussed today, personalized medicine is really the direction in which we're going in prostate cancer care. Genetic and genomic sequencing are increasingly integrated into clinical practice and affecting our treatment decisions.
Where does this put our minority patients? And this again, I'm going to speak predominantly about black Americans, but this affects all Americans of multiple races. We truly are a melting pot and so this really fascinating study by Dan Spratt, who has done actually a large amount of work in this area, and by Felix Feng who presented earlier, they looked at TCGA data for multiple tumor types, 5,729 total samples, and they found in this breakdown that 77% of those samples were from white Americans, 12% black. And you can see other races listed here. The samples were sufficient to detect mutations at a 10% mutational frequency rate only in Caucasian men or white men, but insufficient for all of the other races that I've listed. How do we apply personalized medicine when we don't actually even know if we're identifying mutations that may be driving these patients' cancer?
Dr. Sartor, Oliver Sartor also did a wonderful analysis actually of invitae data, so data from germline sequencing that patients had ordered through routine clinical care. And so he used 3,600 samples and identified actionable mutations across different races. The samples were predominantly from white Americans, 72%, 6% black, 6% Ashkenazi Jewish, and there were few actionable mutations that were identified in the black patients. Although there were many actually identified through the entire population, only 4% of the total actionable mutations were in the black Americans.
It's not just personalized genomic data or genetic data that we don't have. We actually are not necessarily including these patients in our clinical trials, and this was again, work by Dan Spratt. He put together a summary of castration-resistant metastatic prostate cancer trials where he simply listed the total number of patients in all of these landmark trials that led to approvals. And then he showed that only 3% of the patients were actually black Americans. And if you think about the randomization process, only 150 patients out of the 7,200 patients were black and received the active drug. And these were drugs that were clearly active as they're now all approved. Severely limiting our understanding of how these drugs may act specifically in a certain set of patients that we treat every day. But there are questions of whether improving access to these therapies may overcome biological differences should they exist.
Another study by Dan's group and all of his collaborators looked at three different cohorts of patients. There's a SEER population, which is a population that represents general care through insurance in the United States, a Veterans Administration population, which represents patients who all receive similar care through the VA healthcare system. They all have access, they all receive the same care. And then a randomized controlled trial cohort. In the SEER cohort, he found that white men actually had an improved prostate cancer-specific mortality as compared to black men at 10 years with a 0.5% greater prostate cancer-specific mortality for black men.
But when he looked in the VA cohort in which everyone had access to the same care, there was no difference in prostate cancer-specific mortality between black and white men. And when he looked at the randomized controlled trial cohort in which everyone got optimal care, we hope, we found that there was actually an improved mortality. A hazard ratio showing nearly a 20% reduced mortality in black men as compared to white men.
This story I think is fascinating and was similarly represented in some work that Susan Halabi did in a multi, in a meta-analysis of nine studies, including docetaxel or docetaxel containing regimens with 8,800 men with MCRPC. In this study, in this meta-analysis, she found that black men actually were younger, had poor ECOG performance status, higher PSA and testosterone, lower hemoglobin and actually had a similar median overall survival as compared to white men 21 months. But the hazard ratio for mortality in this meta-analysis showed again nearly a 20% reduction in mortality. These patients when given access may actually do better. Though of course there may be some issues of selection and that these patients were in randomized control trials. This is highly encouraging.
Dan George and the group at Duke has done a lot of work suggesting that prospective trials are necessary to really understand biologic differences that may exist and so he's done a nice study called the Abi Race study in which a 100 men, 50 black and 50 white were registered and followed while being treated with abiraterone and they were followed to disease progression.
They found that there was a trend at least towards longer PSA PFS and here you can see that numerically. And additionally, they found differences in toxicity with higher rates of hypertension in black Americans and higher rates of fatigue in Caucasian Americans. Questions about whether these things may be more than simply skin deep.
And he also looked then at the genomics of these patients and the genes may be just really defining and driving both the side effect profiles as we heard when we talked about Asian patients earlier, as well as outcome differences and it's really up to us to continue to hone down our understanding of genetics and genomics, not just in terms of the tumor but the genetics of the patient that may be driving response and the genetics of the patient that may be driving side effects I think are also going to be critically important as we personalize therapy.
Does different biology potentially lead to superior outcomes? We have several examples suggesting that. I mentioned Susan's study. Also, we have David Quinn's study looking at sipuleucel-T, demonstrating an improved mortality for black men. And as well as some studies looking at radiotherapy in localized prostate cancer, also suggesting reduced mortality for black men as compared to Caucasians. We'll see.
There are some efforts that are ongoing. Registry trials that are trying to get more samples and understand not just the United States but the world. The Ironman registry is including 5,000 men, 10 countries, and looking at clinical outcomes, genomics, patient-reported data, really to understand optimal treatment strategies for men with advanced prostate cancer and give us more information, more genetic data so that we can make better decisions. And there's a crowdsourcing effort from the Broad and Dana Farber called Count Me In, where patients, wherever they are across the United States can send in a cheek swab and contribute their genetic data to an information database that can be used by researchers around the world.
In conclusion, disparities in prostate cancer in the US come in many forms just as they do around the world. Both biology and social determinants of health are contributing to these differences. And other disparities, age, geography, education, socioeconomic status, literacy, numeracy exist and are also levers for us to pull in terms of equalizing outcomes for men with prostate cancer. Lack of inclusion in clinical trials in genomic sequencing worsens these disparities and makes it so that when we're treating our patients we don't exactly know what's going to be the outcome and we never know for any individual, but if we have more data we can at least predict better and counsel our patients more appropriately.
Differences in biology may be overcome with improvements in access as we saw and some biologic differences may be associated with improved outcomes for black Americans. And efforts to enhance inclusion in clinical trials and improve access to genomic sequencing are underway with the examples I gave and many others and I look forward to seeing how things change. Issues of disparity must be tackled not just in the US but around the world and I appreciate having the opportunity to speak with everyone when I see that such amazing work is already being done by co-speakers and the interest of the entire community. Thank you very much.
Alicia Morgans: Thank you. And I'm hoping not to hear the cowbell, but I have two small children so I've heard worse. Also, thank you to the organizers for giving me the opportunity to present on the United States and on our disparities. Clearly we can see that around the world there are huge disparities that we all must tackle. This is a small portion of them and I appreciate the time that we're allotted.
This is an outline and I want to say, and we'll discuss that the issues of disparities in the United States are actually bigger than the black and white American disparities that we talk about traditionally and regularly. We'll explore that a little bit. We'll talk about how personalized medicine and the clinical trials that we do can be affected by these disparities and that the data that we have may not be applicable to everyone that we treat. We'll talk about how the access to treatments may be one of the barriers that is definitively impacting these outcomes. And if we improve access, perhaps we can do better. Some ongoing efforts that are trying to make a difference and some conclusions.
I think the most talked about disparity in the United States as you can see here on the incidence curve on the left with the red line showing the incidence in African American men versus other races below and the death curve on the right, demonstrate the black Americans have a two time higher incidence of prostate cancer versus white Americans and two and a half times higher mortality.
And when we think about the causes, I think it's important for us to recognize that it's probably more than just biology. And as I said, I'll show some data to support that, but biology is probably playing some role as well. When we see that the incidence is higher, we don't expect that black men are being screened with PSAs or DREs more commonly than white men. There must be something biologic as well. But the social determinants of health that we see on the right, things like someone's culture, their access to care, financial hardship, their literacy, and numeracy, are actually probably playing a large role in their outcomes as well.
As I mentioned, this is a much bigger story than just black and white. We're really simply scratching the surface when we speak only about these racial issues. There are also disparities by socioeconomic status, age, geography, education and many others and it's also important for us to recognize that disparities exist in therapeutics and in outcomes in terms of disease outcomes but also in supportive care for these patients. And I'll mention that with an example in a few minutes.
On the right is an example of disparities by age that our group put together a few years ago using some SEER data and we saw that as each year went forward, from the 1990s through 2009, there were improvements in terms of prostate cancer-specific survival with each year, but for those patients who were over the age of 75 there was actually no improvement after 1999. This is an example of improvements in survival that are more pronounced for older men. The reasons of which are unclear, but a clear disparity by age.
Another disparity that I'd like to draw our attention to is disparities by education and poverty that can also certainly in the United States, drive a disparity where we have issues of patients not having health insurance and access to healthcare. The Appalachian Mountains or Appalachia is a region that you can see that I've circled here on the right side of the United States. And on the right figure, you can see that these counties are divided up by poverty with red areas demonstrating very desperate poverty and blue areas being areas that have really no economic hardship. In the state of Kentucky, which is circled in that right figure, there's a large area of the Appalachias that are completely economically devastated. As you can see in red.
A group in Kentucky looked at prostate cancer-specific outcomes in men and compared their survival in these men who were from the Appalachian part of Kentucky to those men who were from non-Appalachian Kentucky and found, as you can see in the survival curve on the right, poor survival in men who were from the Appalachian part of the same state. This was associated with lower education attainment, higher poverty and greater number of co-morbid illnesses. Within the same state these issues, and this is a state of predominantly white Americans, these issues, we're driving a very clear disparity.
In California, similar studies were also done. This was a fascinating study done by a group that looked at multiple tumor types and specifically, this is the prostate cancer data. They looked at 270,000 men with prostate cancer and did a mediation analysis to identify which factors were driving disparities by race. If you look at the image on the right, you can see the black Americans were highlighted in blue, white in gray as the reference and black men had poorer survival than other ethnic groups. This was predominantly driven by stage at diagnosis. That was 24% of the disparity indicating again, that I don't think incidence that's higher in black Americans, is actually related to earlier screening. But this was also associated in this analysis with marital status, with neighborhood socioeconomic status and with hospital insurance composition, which really defines whether hospitals are receiving predominantly government supported care or private insurance. And higher rates of things like Medicaid or government supported care were associated with poorer outcomes and we're driving a large part of the disparity in this population in California.
As I mentioned, supportive care also is different or applied differently across patient populations. This is a study that I did with Matthew Smith many years ago when I was one of his fellows and we looked in the SEER Medicare database and found that DEXA screening for bone health, for fragility fractures, was actually quite uncommon for all patients, but was particularly uncommon in patients who were older, over 85, for black Americans and for those patients who had lower educational attainment.
As we know and as we've discussed today, personalized medicine is really the direction in which we're going in prostate cancer care. Genetic and genomic sequencing are increasingly integrated into clinical practice and affecting our treatment decisions.
Where does this put our minority patients? And this again, I'm going to speak predominantly about black Americans, but this affects all Americans of multiple races. We truly are a melting pot and so this really fascinating study by Dan Spratt, who has done actually a large amount of work in this area, and by Felix Feng who presented earlier, they looked at TCGA data for multiple tumor types, 5,729 total samples, and they found in this breakdown that 77% of those samples were from white Americans, 12% black. And you can see other races listed here. The samples were sufficient to detect mutations at a 10% mutational frequency rate only in Caucasian men or white men, but insufficient for all of the other races that I've listed. How do we apply personalized medicine when we don't actually even know if we're identifying mutations that may be driving these patients' cancer?
Dr. Sartor, Oliver Sartor also did a wonderful analysis actually of invitae data, so data from germline sequencing that patients had ordered through routine clinical care. And so he used 3,600 samples and identified actionable mutations across different races. The samples were predominantly from white Americans, 72%, 6% black, 6% Ashkenazi Jewish, and there were few actionable mutations that were identified in the black patients. Although there were many actually identified through the entire population, only 4% of the total actionable mutations were in the black Americans.
It's not just personalized genomic data or genetic data that we don't have. We actually are not necessarily including these patients in our clinical trials, and this was again, work by Dan Spratt. He put together a summary of castration-resistant metastatic prostate cancer trials where he simply listed the total number of patients in all of these landmark trials that led to approvals. And then he showed that only 3% of the patients were actually black Americans. And if you think about the randomization process, only 150 patients out of the 7,200 patients were black and received the active drug. And these were drugs that were clearly active as they're now all approved. Severely limiting our understanding of how these drugs may act specifically in a certain set of patients that we treat every day. But there are questions of whether improving access to these therapies may overcome biological differences should they exist.
Another study by Dan's group and all of his collaborators looked at three different cohorts of patients. There's a SEER population, which is a population that represents general care through insurance in the United States, a Veterans Administration population, which represents patients who all receive similar care through the VA healthcare system. They all have access, they all receive the same care. And then a randomized controlled trial cohort. In the SEER cohort, he found that white men actually had an improved prostate cancer-specific mortality as compared to black men at 10 years with a 0.5% greater prostate cancer-specific mortality for black men.
But when he looked in the VA cohort in which everyone had access to the same care, there was no difference in prostate cancer-specific mortality between black and white men. And when he looked at the randomized controlled trial cohort in which everyone got optimal care, we hope, we found that there was actually an improved mortality. A hazard ratio showing nearly a 20% reduced mortality in black men as compared to white men.
This story I think is fascinating and was similarly represented in some work that Susan Halabi did in a multi, in a meta-analysis of nine studies, including docetaxel or docetaxel containing regimens with 8,800 men with MCRPC. In this study, in this meta-analysis, she found that black men actually were younger, had poor ECOG performance status, higher PSA and testosterone, lower hemoglobin and actually had a similar median overall survival as compared to white men 21 months. But the hazard ratio for mortality in this meta-analysis showed again nearly a 20% reduction in mortality. These patients when given access may actually do better. Though of course there may be some issues of selection and that these patients were in randomized control trials. This is highly encouraging.
Dan George and the group at Duke has done a lot of work suggesting that prospective trials are necessary to really understand biologic differences that may exist and so he's done a nice study called the Abi Race study in which a 100 men, 50 black and 50 white were registered and followed while being treated with abiraterone and they were followed to disease progression.
They found that there was a trend at least towards longer PSA PFS and here you can see that numerically. And additionally, they found differences in toxicity with higher rates of hypertension in black Americans and higher rates of fatigue in Caucasian Americans. Questions about whether these things may be more than simply skin deep.
And he also looked then at the genomics of these patients and the genes may be just really defining and driving both the side effect profiles as we heard when we talked about Asian patients earlier, as well as outcome differences and it's really up to us to continue to hone down our understanding of genetics and genomics, not just in terms of the tumor but the genetics of the patient that may be driving response and the genetics of the patient that may be driving side effects I think are also going to be critically important as we personalize therapy.
Does different biology potentially lead to superior outcomes? We have several examples suggesting that. I mentioned Susan's study. Also, we have David Quinn's study looking at sipuleucel-T, demonstrating an improved mortality for black men. And as well as some studies looking at radiotherapy in localized prostate cancer, also suggesting reduced mortality for black men as compared to Caucasians. We'll see.
There are some efforts that are ongoing. Registry trials that are trying to get more samples and understand not just the United States but the world. The Ironman registry is including 5,000 men, 10 countries, and looking at clinical outcomes, genomics, patient-reported data, really to understand optimal treatment strategies for men with advanced prostate cancer and give us more information, more genetic data so that we can make better decisions. And there's a crowdsourcing effort from the Broad and Dana Farber called Count Me In, where patients, wherever they are across the United States can send in a cheek swab and contribute their genetic data to an information database that can be used by researchers around the world.
In conclusion, disparities in prostate cancer in the US come in many forms just as they do around the world. Both biology and social determinants of health are contributing to these differences. And other disparities, age, geography, education, socioeconomic status, literacy, numeracy exist and are also levers for us to pull in terms of equalizing outcomes for men with prostate cancer. Lack of inclusion in clinical trials in genomic sequencing worsens these disparities and makes it so that when we're treating our patients we don't exactly know what's going to be the outcome and we never know for any individual, but if we have more data we can at least predict better and counsel our patients more appropriately.
Differences in biology may be overcome with improvements in access as we saw and some biologic differences may be associated with improved outcomes for black Americans. And efforts to enhance inclusion in clinical trials and improve access to genomic sequencing are underway with the examples I gave and many others and I look forward to seeing how things change. Issues of disparity must be tackled not just in the US but around the world and I appreciate having the opportunity to speak with everyone when I see that such amazing work is already being done by co-speakers and the interest of the entire community. Thank you very much.