Harm-to-Benefit of Three Decades of Prostate Cancer Screening in Black Men - Jonathan Shoag
January 31, 2023
In a discussion between Alicia Morgans and Jonathan Shoag, the importance of PSA screening for prostate cancer is emphasized, particularly considering its controversial history. Introduced over 30 years ago, PSA screening initially led to an increased incidence of prostate cancer diagnoses without a corresponding decrease in mortality. This led to skepticism and a backlash against the procedure, resulting in changes to screening guidelines. Dr. Shoag details two significant trials on PSA screening and the subsequent evolution in understanding that has shown PSA screening's actual benefits in reducing mortality. Emphasizing the greater potential benefits for Black men, who have higher risks of prostate cancer, the conversation also covers the need to update guidelines and screening practices, including new findings that longer follow-up times demonstrate greater benefits of PSA screening.
Biographies:
Jonathan Shoag, MD, Urologist, University Hospitals Cleveland, Assistant Professor of Urology at Case Western Reserve University, Case Comprehensive Cancer Center, Cleveland, OH
Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, MA
Biographies:
Jonathan Shoag, MD, Urologist, University Hospitals Cleveland, Assistant Professor of Urology at Case Western Reserve University, Case Comprehensive Cancer Center, Cleveland, OH
Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, MA
Read the Full Video Transcript
Alicia Morgans: Hi, my name is Alicia Morgans. I'm a GU Medical Oncologist at Dana-Farber Cancer Institute. I'm so excited to talk today with Dr. Jonathan Shoag, who is an Assistant Professor of Urology at Case Western Reserve University School of Medicine in Cleveland, as well as being a urologist at University Hospitals, Cleveland Medical Center. Thank you so much for being here with me today, Dr. Shoag.
Jonathan Shoag: Thank you so much for having me. I really appreciate it.
Alicia Morgans: Wonderful. Well, I appreciate you, and I really appreciate you emphasizing for our community and for the listeners, the importance of PSA screening and helping us to understand where we've been and where we're going with that particular approach to diagnosing and following prostate cancer. Let's focus for a minute on the background. Why is PSA screening important and where have we been in the past in terms of this screening technique?
Jonathan Shoag: Sure. So PSA screening has been controversial since we started doing it over 30 years ago. The PSA blood test was introduced in the late '80s, early '90s, as a way to diagnose prostate cancer early and was very rapidly taken up at a population level. What happened was we saw the incidence of prostate cancer go up really dramatically without much change in mortality throughout the '90s, and even, in the early 2000s. In that setting, people became very skeptical of PSA screening, because we were finding all these prostate cancers and mortality at a population level hadn't really changed.
That coincided with the publication of two large randomized trials on PSA screening. The PLCO trial done in the United States and the ERSPC trial done in Europe. The PLCO trial, which was done in the US, was reported initially around nine years of follow up, in the same issue as the New England Journal of Medicine in 2009 as the ERSPC, and the PLCO trial showed no difference between screening control arms in prostate cancer mortality, while the ERSPC trial showed about a 20% relative risk reduction. That 20% relative risk reduction was actually a very small absolute magnitude of reduction, and so what came out of that, the takeaway bullet point was that, for 1000 men screened, we prevented one man from dying of prostate cancer, if that was even true, because the PLCO trial showed no benefit.
In that setting, PSA screening became the villain of the less is more movement. We were finding all this prostate cancer, we weren't helping anyone, and a lot of major organizations, including the United States Preventive Services Task Force, but also the AFP, came out against screening, very vocally, and even our own organizations like the AUA walked back their recommendations for screening advocating, shared decision making, etc.
Since then, and this is in the early 2010s, 2012 was when this came to a head and actually, many of us don't remember, but Medicare was actually evaluating an initiative to disincentivize PSA screening, so that was in the conversation is that you'd get primary care doctors actually penalized for performing PSA screening. Work we did in 2016 actually showed that the PLCO trial was much more contaminated than most people thought, meaning essentially the entire control arm had PSA screening, and more and more epidemiologic evidence came out, showing that PSA screening actually really did reduce mortality at a population level, that stopping screening seemed to be increasing mortality.
Concurring with that, the ERSPC trial with longer follow-up has shown larger and larger absolute benefits from screening, so as follow-up gets longer, we prevent more men from dying. That 20% relative risk reduction has been preserved, but more men are dying of prostate cancer. The United States Preventive Services Task Force changed their guidelines and actually now doesn't incorporate the PLCO trial very much in making their decisions, and they changed it from a D to a C. So that's something that, obviously, we're very happy about, but the current nomograms, and what most doctors use who would be doing PSA screening, still are pretty unfavorable towards screening.
The current task force infographic shows that for a thousand men screened, we over treat 80, causing impotence and incontinence at very, very high levels to maybe prevent one mortality. In generating these figures, which is true of the current Cochrane review, the current AFP guidelines, people have used around a 10 to 13-year window of looking at the benefits of screening. So we had the hypothesis that if we look over a longer period of time, the benefits of screening will be larger, particularly as compared to the harm.
If you can imagine a screening test that finds a cancer early and you look at one year outcomes, you would see a very, very high incidence, but you wouldn't have prevented any mortality because no one's died of the cancer yet. So we know from natural history studies that prostate cancer has a very, very long natural history, and even clinically detected prostate cancer, men don't usually die of that. It's all 15 years after diagnosis. Intuitively, that makes sense too. We find prostate cancer in men in their '50s and '60s and '70s, and men don't die of prostate cancer until their '80s usually. So under that idea, that follow time really matters, we use the epidemiologic evidence to say, "Well, it's now been 30 years since we started doing PSA screening, what can we tell from this data to measure the harms and benefits?"
Another advantage of this is the epidemiologic data allows us to look specifically at Black men, who aren't included in clinical trials of PSA screening that we have existed. The reason to look at Black men specifically is that they have a higher risk of prostate cancer incidence, as well as prostate cancer mortality. Because of that, their potential benefit from screening is substantially greater.
We had a similar article come out, around two years ago now, where we used the randomized data and modeled it going long term, and that was also published in the New England Journal of Medicine, and we showed that just taking longer follow-up or modeling out longer follow up results, it results in dramatically more favorable numbers for screening. So using the epidemiologic data, we're now able to take the same idea and model that specifically for Black men, as well as all men. So that was the impetus for this current study.
Alicia Morgans: Yeah, to your point, it's so important I think that, even the USPSTF as it was revising its grading and recommendation over the years really acknowledged and said that there are populations where we need more data, particularly in Black men, and so there has been a call to action. So it's, I think, really important that your team has been able to look at this specifically and try to understand, in this population, at highest risk, and with the most to gain, perhaps, given their mortality risk, what that benefit is actually going to be.
I think it's also important for us all to acknowledge that mortality may not be the only important endpoint in PSA screening, and of course, there are other benefits like prevention of suffering and complications, even prevention of long term treatment with ADT that can be helpful for patients. So there are a lot of issues around all of this and so really important work that you're doing. Can you tell us, what did you find in this study that you most recently did?
Jonathan Shoag: Correct. Yeah, a hundred percent agree. We were able to look at mortality because that's the data we had. The development of metastasis is not captured in CR and the types of studies we used. What we did is we looked at two approaches to measure the harms of screening, which we called the excess incidence. How many more men were diagnosed with prostate cancer as a consequence of screening, who wouldn't be diagnosed otherwise?
We used a very simple approach, where we said, "Well, what was the incidence before screening? Everyone who was diagnosed extra compared to that was overdiagnosed," or we used a model which was performed by our colleagues at Fred Hutchinson Cancer Center to model out how many extra men were diagnosed. That's the numerator of the equation, the number needed to treat, the number needed to diagnose to prevent a death. The denominator of that equation is how many deaths did not occur since the initiation of screening, and we can make different assumptions about the magnitude of that denominator.
We can assume that the entire mortality reduction since the introduction of screening was the consequence of screening, and there's been about a 50% reduction in mortality. We can say that half of that is due to screening and the rest is due to improvements in treatment and care. Or we can say a smaller fraction is due to that, and so there's a bunch of different assumptions we can make as that, and that's a point of debate. Conservatively, we said it's about half. And the modeling studies would support that being the correct number. That screening accounts for about half of the reduction in mortality since it's advent, but we modeled it out using different assumptions.
What we found is that, for all men, the numbers are remarkably more favorable when we look over a longer term than any of the current guidelines really use. For men of all races, the number needed to treat, the number needed to diagnose are around 10, if not lower, and again, dependent on the assumptions. There's a paper from Gilbert Welch and Pete Halvorsen that was published using much shorter follow up and they calculated the number needed to treat in the twenties, suggesting that PSA screening wasn't beneficial. Just the counter for longer follow-up using that exact same methodology, we get a number of six. So just the longer follow-up alone really reduces that number needed to treat, to prevent one prostate cancer death.
When we look specifically at Black men, those numbers are remarkably low, so depending on the assumptions again, to the low single digits. Comparing that to existing infographics and guidelines and recommendations is pretty striking, so it's very hard to argue that a number needed to treat of five or two, what doesn't justify PSA too. Our numbers are really robust to the assumptions that we use and we include multiple different assumptions, as well as sensitivity analyses. I find them very compelling. That PSA screening, in my opinion, for all men, but certainly for men at higher risk, is really beneficial.
Just to contextualize that number needed to treat, because it's hard to... If you're thinking, "Well, I have to treat four men to prevent one death," but the numbers from PROTRACT would suggest that the number needed to treat to cause one case of erectile dysfunction or one case of urinary incontinence from our treatment are around the same order of magnitude. The numbers for protect to cause one case of erectile dysfunction, I think are four for surgery and eight for radiation, and that's around the same order of magnitude as we're seeing for mortality benefit.
If you think of those trade-offs, I have a really hard time suggesting that all men shouldn't be screened for prostate cancer, particularly those at higher risk. There's obviously different ways to screen and there's different screening algorithms and where active surveillance and MRI are all very important if we're initiating a nationwide screening program, but some PSA based screening, I think, should be a part of every healthy man's healthcare at some point.
Alicia Morgans: So, obviously I'm biased, I've been doing this for a long time and I actually, I've been practicing through all of the changes with the USPSTF and I witnessed it and I also witnessed patients coming in who say, "My doctor just stopped screening and didn't tell me," and of course, when they come to the medical oncologist, in many cases, they have either very locally advanced disease or metastatic disease. It is definitely, it's heartbreaking, especially when patients say, "But I just didn't know." So I'm glad that we're informing patients or have been trying to inform patients through things like infographics, though clearly, the data that maybe we're sharing needs to be updated. Do you and the team have thoughts about getting this data into the hands of those who may have the ability to help guide our patients more effectively?
Jonathan Shoag: Yeah, so actually, in that regard, our article this week came out, also with a patient platform of a patient telling that same story. Young African American man diagnosed with de novo metastatic disease, wasn't screened, actually had a family history and certainly in my practice, I see this more and more. It sounds like you're seeing it as well. Also, epidemiologically, we're seeing it. We're able to see it in SEER and that's a pretty striking thing that we're able to see these differences, and the patient in this patient platform is 56.
We've had tremendous advances in the treatment of metastatic and locally advanced disease, but better to prevent this from happening in the first place. With a cheap, easy intervention, that's very widely deployable and had been deployed before, that we'd walked back, it's pretty, pretty shocking that we're not doing that anymore.
Then in terms of how to get this into the right hands, I think doing... This is my hope, is to have these conversations and obviously, try to meet with anyone who will listen and appreciate the opportunity to talk about it here, because hopefully, we can get the word out a little bit better.
Alicia Morgans: Thank you so much for taking the time to go through this with us. And, of course, thank you for doing this work, because what was clear from the USPSTF statement is that there is a call for more investigation. That more research needs to be done, and only with research, can that group reevaluate the data over time and really, I think, hopefully adjust the recommendation and the guidance to primary care teams and to patients to raise awareness that this may be a more effective strategy than we previously thought and may have fewer harms than we previously thought. And that risk benefit ratio may really be more appealing to a majority of patients than we thought. So thank you for doing the work and thank you for sharing your perspective with us today. I really appreciate your time.
Jonathan Shoag: Thanks so much for having me.
Alicia Morgans: Hi, my name is Alicia Morgans. I'm a GU Medical Oncologist at Dana-Farber Cancer Institute. I'm so excited to talk today with Dr. Jonathan Shoag, who is an Assistant Professor of Urology at Case Western Reserve University School of Medicine in Cleveland, as well as being a urologist at University Hospitals, Cleveland Medical Center. Thank you so much for being here with me today, Dr. Shoag.
Jonathan Shoag: Thank you so much for having me. I really appreciate it.
Alicia Morgans: Wonderful. Well, I appreciate you, and I really appreciate you emphasizing for our community and for the listeners, the importance of PSA screening and helping us to understand where we've been and where we're going with that particular approach to diagnosing and following prostate cancer. Let's focus for a minute on the background. Why is PSA screening important and where have we been in the past in terms of this screening technique?
Jonathan Shoag: Sure. So PSA screening has been controversial since we started doing it over 30 years ago. The PSA blood test was introduced in the late '80s, early '90s, as a way to diagnose prostate cancer early and was very rapidly taken up at a population level. What happened was we saw the incidence of prostate cancer go up really dramatically without much change in mortality throughout the '90s, and even, in the early 2000s. In that setting, people became very skeptical of PSA screening, because we were finding all these prostate cancers and mortality at a population level hadn't really changed.
That coincided with the publication of two large randomized trials on PSA screening. The PLCO trial done in the United States and the ERSPC trial done in Europe. The PLCO trial, which was done in the US, was reported initially around nine years of follow up, in the same issue as the New England Journal of Medicine in 2009 as the ERSPC, and the PLCO trial showed no difference between screening control arms in prostate cancer mortality, while the ERSPC trial showed about a 20% relative risk reduction. That 20% relative risk reduction was actually a very small absolute magnitude of reduction, and so what came out of that, the takeaway bullet point was that, for 1000 men screened, we prevented one man from dying of prostate cancer, if that was even true, because the PLCO trial showed no benefit.
In that setting, PSA screening became the villain of the less is more movement. We were finding all this prostate cancer, we weren't helping anyone, and a lot of major organizations, including the United States Preventive Services Task Force, but also the AFP, came out against screening, very vocally, and even our own organizations like the AUA walked back their recommendations for screening advocating, shared decision making, etc.
Since then, and this is in the early 2010s, 2012 was when this came to a head and actually, many of us don't remember, but Medicare was actually evaluating an initiative to disincentivize PSA screening, so that was in the conversation is that you'd get primary care doctors actually penalized for performing PSA screening. Work we did in 2016 actually showed that the PLCO trial was much more contaminated than most people thought, meaning essentially the entire control arm had PSA screening, and more and more epidemiologic evidence came out, showing that PSA screening actually really did reduce mortality at a population level, that stopping screening seemed to be increasing mortality.
Concurring with that, the ERSPC trial with longer follow-up has shown larger and larger absolute benefits from screening, so as follow-up gets longer, we prevent more men from dying. That 20% relative risk reduction has been preserved, but more men are dying of prostate cancer. The United States Preventive Services Task Force changed their guidelines and actually now doesn't incorporate the PLCO trial very much in making their decisions, and they changed it from a D to a C. So that's something that, obviously, we're very happy about, but the current nomograms, and what most doctors use who would be doing PSA screening, still are pretty unfavorable towards screening.
The current task force infographic shows that for a thousand men screened, we over treat 80, causing impotence and incontinence at very, very high levels to maybe prevent one mortality. In generating these figures, which is true of the current Cochrane review, the current AFP guidelines, people have used around a 10 to 13-year window of looking at the benefits of screening. So we had the hypothesis that if we look over a longer period of time, the benefits of screening will be larger, particularly as compared to the harm.
If you can imagine a screening test that finds a cancer early and you look at one year outcomes, you would see a very, very high incidence, but you wouldn't have prevented any mortality because no one's died of the cancer yet. So we know from natural history studies that prostate cancer has a very, very long natural history, and even clinically detected prostate cancer, men don't usually die of that. It's all 15 years after diagnosis. Intuitively, that makes sense too. We find prostate cancer in men in their '50s and '60s and '70s, and men don't die of prostate cancer until their '80s usually. So under that idea, that follow time really matters, we use the epidemiologic evidence to say, "Well, it's now been 30 years since we started doing PSA screening, what can we tell from this data to measure the harms and benefits?"
Another advantage of this is the epidemiologic data allows us to look specifically at Black men, who aren't included in clinical trials of PSA screening that we have existed. The reason to look at Black men specifically is that they have a higher risk of prostate cancer incidence, as well as prostate cancer mortality. Because of that, their potential benefit from screening is substantially greater.
We had a similar article come out, around two years ago now, where we used the randomized data and modeled it going long term, and that was also published in the New England Journal of Medicine, and we showed that just taking longer follow-up or modeling out longer follow up results, it results in dramatically more favorable numbers for screening. So using the epidemiologic data, we're now able to take the same idea and model that specifically for Black men, as well as all men. So that was the impetus for this current study.
Alicia Morgans: Yeah, to your point, it's so important I think that, even the USPSTF as it was revising its grading and recommendation over the years really acknowledged and said that there are populations where we need more data, particularly in Black men, and so there has been a call to action. So it's, I think, really important that your team has been able to look at this specifically and try to understand, in this population, at highest risk, and with the most to gain, perhaps, given their mortality risk, what that benefit is actually going to be.
I think it's also important for us all to acknowledge that mortality may not be the only important endpoint in PSA screening, and of course, there are other benefits like prevention of suffering and complications, even prevention of long term treatment with ADT that can be helpful for patients. So there are a lot of issues around all of this and so really important work that you're doing. Can you tell us, what did you find in this study that you most recently did?
Jonathan Shoag: Correct. Yeah, a hundred percent agree. We were able to look at mortality because that's the data we had. The development of metastasis is not captured in CR and the types of studies we used. What we did is we looked at two approaches to measure the harms of screening, which we called the excess incidence. How many more men were diagnosed with prostate cancer as a consequence of screening, who wouldn't be diagnosed otherwise?
We used a very simple approach, where we said, "Well, what was the incidence before screening? Everyone who was diagnosed extra compared to that was overdiagnosed," or we used a model which was performed by our colleagues at Fred Hutchinson Cancer Center to model out how many extra men were diagnosed. That's the numerator of the equation, the number needed to treat, the number needed to diagnose to prevent a death. The denominator of that equation is how many deaths did not occur since the initiation of screening, and we can make different assumptions about the magnitude of that denominator.
We can assume that the entire mortality reduction since the introduction of screening was the consequence of screening, and there's been about a 50% reduction in mortality. We can say that half of that is due to screening and the rest is due to improvements in treatment and care. Or we can say a smaller fraction is due to that, and so there's a bunch of different assumptions we can make as that, and that's a point of debate. Conservatively, we said it's about half. And the modeling studies would support that being the correct number. That screening accounts for about half of the reduction in mortality since it's advent, but we modeled it out using different assumptions.
What we found is that, for all men, the numbers are remarkably more favorable when we look over a longer term than any of the current guidelines really use. For men of all races, the number needed to treat, the number needed to diagnose are around 10, if not lower, and again, dependent on the assumptions. There's a paper from Gilbert Welch and Pete Halvorsen that was published using much shorter follow up and they calculated the number needed to treat in the twenties, suggesting that PSA screening wasn't beneficial. Just the counter for longer follow-up using that exact same methodology, we get a number of six. So just the longer follow-up alone really reduces that number needed to treat, to prevent one prostate cancer death.
When we look specifically at Black men, those numbers are remarkably low, so depending on the assumptions again, to the low single digits. Comparing that to existing infographics and guidelines and recommendations is pretty striking, so it's very hard to argue that a number needed to treat of five or two, what doesn't justify PSA too. Our numbers are really robust to the assumptions that we use and we include multiple different assumptions, as well as sensitivity analyses. I find them very compelling. That PSA screening, in my opinion, for all men, but certainly for men at higher risk, is really beneficial.
Just to contextualize that number needed to treat, because it's hard to... If you're thinking, "Well, I have to treat four men to prevent one death," but the numbers from PROTRACT would suggest that the number needed to treat to cause one case of erectile dysfunction or one case of urinary incontinence from our treatment are around the same order of magnitude. The numbers for protect to cause one case of erectile dysfunction, I think are four for surgery and eight for radiation, and that's around the same order of magnitude as we're seeing for mortality benefit.
If you think of those trade-offs, I have a really hard time suggesting that all men shouldn't be screened for prostate cancer, particularly those at higher risk. There's obviously different ways to screen and there's different screening algorithms and where active surveillance and MRI are all very important if we're initiating a nationwide screening program, but some PSA based screening, I think, should be a part of every healthy man's healthcare at some point.
Alicia Morgans: So, obviously I'm biased, I've been doing this for a long time and I actually, I've been practicing through all of the changes with the USPSTF and I witnessed it and I also witnessed patients coming in who say, "My doctor just stopped screening and didn't tell me," and of course, when they come to the medical oncologist, in many cases, they have either very locally advanced disease or metastatic disease. It is definitely, it's heartbreaking, especially when patients say, "But I just didn't know." So I'm glad that we're informing patients or have been trying to inform patients through things like infographics, though clearly, the data that maybe we're sharing needs to be updated. Do you and the team have thoughts about getting this data into the hands of those who may have the ability to help guide our patients more effectively?
Jonathan Shoag: Yeah, so actually, in that regard, our article this week came out, also with a patient platform of a patient telling that same story. Young African American man diagnosed with de novo metastatic disease, wasn't screened, actually had a family history and certainly in my practice, I see this more and more. It sounds like you're seeing it as well. Also, epidemiologically, we're seeing it. We're able to see it in SEER and that's a pretty striking thing that we're able to see these differences, and the patient in this patient platform is 56.
We've had tremendous advances in the treatment of metastatic and locally advanced disease, but better to prevent this from happening in the first place. With a cheap, easy intervention, that's very widely deployable and had been deployed before, that we'd walked back, it's pretty, pretty shocking that we're not doing that anymore.
Then in terms of how to get this into the right hands, I think doing... This is my hope, is to have these conversations and obviously, try to meet with anyone who will listen and appreciate the opportunity to talk about it here, because hopefully, we can get the word out a little bit better.
Alicia Morgans: Thank you so much for taking the time to go through this with us. And, of course, thank you for doing this work, because what was clear from the USPSTF statement is that there is a call for more investigation. That more research needs to be done, and only with research, can that group reevaluate the data over time and really, I think, hopefully adjust the recommendation and the guidance to primary care teams and to patients to raise awareness that this may be a more effective strategy than we previously thought and may have fewer harms than we previously thought. And that risk benefit ratio may really be more appealing to a majority of patients than we thought. So thank you for doing the work and thank you for sharing your perspective with us today. I really appreciate your time.
Jonathan Shoag: Thanks so much for having me.