Prostate Cancer Screening is a Shared Decision - Matthew Cooperberg

September 10, 2018

Alicia Morgans and Matt Cooperberg discuss the controversial topic of how to optimally screen for prostate cancer based on the final recommendations from the U.S. Preventive Services Task Force. 


Matthew Cooperberg, MD, MPH, FACS, UCSF Helen Diller Family Comprehensive Cancer Center

Alicia Morgans, MD, MPH

Read the full video transcript

Dr. Alicia Morgans: Hi and welcome to more of our AUA 2018 coverage. We're delighted to be able to speak with Dr. Matt Cooperberg, who is an associate professor of Urology, epidemiology, and biostatistics at the University of California San Francisco. Great to see you, Matt. 

Dr. Matt Cooperberg: Great to be here. I'm excited.

Dr. Alicia Morgans: So, I think really one of the things that is most exciting in terms of localized prostate cancer and diagnosis of prostate cancer in general that just came out last week in JAMA are the final recommendations for the USPSTF on PSA screening guidelines. I'd love to hear your thoughts on the changes that were just published. 

Dr. Matt Cooperberg: Sure. So, it's a perennially controversial question, of course, how we should be optimally screening for prostate cancer. The new C recommendation is not much of a change from the draft that they released late last year. It's obviously step in the right direction. They've gone from recommending that nobody be screened ever, to a screening recommendation for men 55 to 69, that they at least have a shared decision-making conversation with their primary care doctors about whether or not to screen.

With this change, they are now much more in line with most of the other guidelines out there. The AUA, the ACS, the NCCN have all been recommending shared decision making for a number of years. So, the task force is no longer the outlier in that regard. It is progress. There are still some major flaws in the way they program their evidence review, in the way they handled the data from a methodological standpoint and in some of the specifics of the conclusion. But like I said, it's a step in the right direction. 

I think that one of the major problems is still this 55-start age for screening, it is probably later than it should be. There's actually quite a bit of data that earlier screening is more effective because there's less BPH, so less contamination of PSA, if you will, in younger men. And it's actually quite a good evidence base for screening men in the 50 to 55 age range, based on studies and outside of the core age of the ERSPC, we've got randomized trial data and then we've got really good cohort studies from Malmo and from HPFS looking at even earlier ages.

I think they could have made a very defensible argument about screening 50 to 69 based on those RCTs and they chose not to do that, which is, which is unfortunate. But again, at least now there's concordance across the guidelines for shared decision making. 

Then the question becomes what should that shared decision-making conversation look like? Some of the other limitations in the statement is a myopic view of screening benefits at 13 years. The screening discussion for men 50 or even 60 or frankly these days even 70 is not about a 13-year horizon. It's about a 20, 30-year horizon. 

When you look at the patient guide that kind of came with the guideline, it's very misleading because they talk about one life saved at 13 years. Well, that's not the relevant timeframe that a man cares about who is 50 or 60, it really is the 20, 25 years. And over that timeframe, the number of lives saved with screening of course increases substantially. 

The harms are also overstated. They take sort of a worst-case scenario in terms of the risks of incontinence and long-term sexual dysfunction after treatment. You know, they cite studies that are in the literature, they are published, but it's a selected view of the literature, and they ignore a lot of studies showing the better outcomes that are achievable in higher volume centers. 

So, it's definitely progress, it's definitely step in the right direction. I would still always encourage primary care doctors to learn the evidence a little bit themselves to look at the other guidelines. I still think the NCCN is probably the closest to what we should be doing. 

The problem with all the guidelines though is that none of them really tells us how to screen, in part because the RCTs have not really addressed those questions. We know from PLCO that annual screening starting at 55 or 60 or 65 is not optimal. 

We have a good sense from some of these cohort studies that men who have a low baseline test really can defer further testing for many years. And I think that's the concept that we're really trying to push now, is this idea of smarter screening. We should be thinking about PSA more like a colonoscopy. If your PSA is low at baseline, and low means less than one for young men, not four, you really don't need another test for a number of years. 

For men who have a slightly high PSA, we're now very interested in secondary testing with things like 4K and PHI or select MDx, which are these secondary tests with very good negative predictive value for clinically significant prostate cancer. MRI is also in this category.

An elevated PSA does not mean a biopsy. And then of course, biopsy and diagnosis of prostate cancer should not and cannot mean immediate treatment, and that's where I think risk stratification is finally getting out of the academic world into real world practice, when we look at what's happening with active surveillance rates around the country. 

Dr. Alicia Morgans: So, can you speak a bit more about that? I know that active surveillance is really part of your passion and the USPSTF's guidelines, I think, were both meant to enhance our ability to detect these higher risk cancers, but also make sure that we're not over-treating patients. And I think that you as an individual and your field of urology has really pivoted on practices in the 90s where everyone with a diagnosis of prostate cancer seemed to get a prostatectomy, and active surveillance really is limiting harms to patients and is increasing in its frequency of use. So, I'd love to hear your thoughts on it. 

Dr. Matt Cooperberg: Yeah, that's exactly right. The task force, it's all about benefits and harms and everybody acknowledges that there are benefits of screening that are harms of screening. The question is how do you evaluate those and how do you weight them? And there is no question that, of course, the most important harm of screening is the harm of overdiagnosis and overtreatment of low risk disease. 

We know low risk prostate cancer kills almost nobody, even over 20, 30 years follow up. There are those of us who think we shouldn't even call it cancer when we're talking about this one microfocus of low-grade disease and we do a great disservice to men by giving them that label, that diagnosis. But active surveillance really has, I think finally gotten out of the academic world, you're absolutely right, through the 90s and even through the first decade of this century, a diagnosis of prostate cancer meant treatment for prostate cancer. 

And we were among the centers that showed that over and over again in the community-based registries that we've run. Really, just in the current decade, we really have seen a paradigm change, where whereas before active surveillance was used for less than 10% of men now is 40% in the large registry that we find here. In the Michigan Consortium, which is collecting data across the entire state, it's up to 50%. 

I was actually just at one of the poster sessions here just a few minutes ago and there are data from the Optum database, which is a big national payer database looking at both Medicare Advantage patients and younger men with private pay insurance where surveillance is up to about 20% of all localized prostate cancer, not just low risk disease. And if you do the math, that is again going to be about 40 - 50% of low risk prostate cancer patients. 

We really are seeing a real paradigm change. It's amazing how quickly it's happened. It really is just in this current decade. There are those of us that think part of this probably was a reaction to the taskforce, and if the taskforce did anything good with the D recommendation, I think it was really forcing this wakeup call that over treating prostate cancer is not an acceptable stance for us to have as a treating community. I was definitely in the camp that was saying for years, if we don't fix the over treatment problem, we cannot regain control of the conversation about early detection. 

Dr. Alicia Morgans: Yeah, absolutely, and it's wonderful. And I just want to state that again, 40 - 50%, potentially, of these low risk patients who are appropriate for active surveillance are now there. And I agree with you, that's really fast. Medicine does not usually move very quickly. So, phenomenal work by your group and others at raising awareness and helping providers and patients feel safe in that decision. 

You know, one of the ways that people can feel safe in that decision is to use potentially molecular or genomic data to help risk stratify and help them figure out is this localized disease one that's going to ultimately be a bad one, or is this maybe a lower risk localized disease? Have you integrated these risk stratification methods into your practice? 

Dr. Matt Cooperberg: Yeah. So, the short answer is yes, the long answer is it's actually really complicated how to do that optimally. So yeah, what it means to be on active surveillance is evolving quite quickly. First of all, the 40% number is actually still probably quite a bit too low if you look at Sweden and other places up to 80 percent. And that's probably where we should be in terms of proportion of low risk men who at least start on surveillance. 

But surveillance is not completely benign. It does entail serial biopsies and the biopsies are a critical part of the surveillance paradigm. If you look at data like those from the ProtecT trial out of the UK, if you have a monitoring strategy that does not include the biopsies, you will definitely miss progressive cases. And I think that's what the early results from ProtecT have shown. 

There are risks to biopsy, there's discomfort from biopsy, and we're really interested in modulating the surveillance paradigm. So, the first question is who should be on surveillance? And then once we get men on surveillance, how can we do it more safely? How can we avoid harm for men who don't need such frequent intervention, and how can we avoid missing cases that actually are going to progress more quickly? 

So, genomics will play a role. I think trying to pinpoint exactly what that role is challenging. We have a number of markers that have now reached the clinical marketplace in the US. There's the Polaris test, the Oncotype and the Decipher, which are all tissue-based tests. We have blood and urine tests which are trying to enter the post-diagnosis space, and then we have MRI, and MRI really needs to play in this same space and answer the same questions about proving efficacy. 

I think there was this New England Journal paper about using MRI in the pre-diagnosis space as a secondary test between PSA screening and biopsy. In the post-diagnosis space, there's a lot of interest in MRI as in biomarkers. But again, the devil is in the details. We know that all these tests, the genomics and MR can improve prognosis. In other words, they can improve on Gleason score, PSA stage, extent of biopsy involvement, maybe even PSA density in terms of predicting who's going to do badly, who's going to do well. 

But they have not gotten to the point of being able to really directly guide a clinical decision at the point of care in a way that is easy enough and reproducible enough for the typical patient and typical community urologist to implement in a kind of straightforward, reproducible way that everybody's looking for. 

To put it more bluntly and everybody wants a binary test. The joke I always use is everyone wants a pregnancy test, and if the test turns blue, you take out the prostate. It's never really going to be that easy because prostate cancer is not a binary condition, and it's not as easy as low risk versus high risk. It's a continuum of risk. 

What the tests tell us is, we know this is not a black or white question. It's a shade of gray, and maybe it tells a slightly lighter or slightly darker shade of gray. But it can't tell us what to do, at least not yet. None of the tests can quite get to the point of a binary answer. So, for that reason, I think the way they really get used is still a very individualized and very customized. At UCSF and I think a lot of other academic centers, we're really using them to try to help push the envelope in terms of who is the surveillance candidate. 

I think the value is not necessarily there for men who have a low volume, Gleason six cancer, it's for those who have the higher volume tumors or those with low volume Gleason three plus fours for whom surveillance is not the obvious decision but may be a reasonable decision. I think, in those cases the markers can be very useful as tiebreakers. 

But the other place I'm really optimistic for markers, and this is where we are just getting studies off the ground now, is can we really use them to modulate surveillance. So, if you think about every hundred men that we identify as surveillance candidates, we know that about half of them are going to come to treatment in the first five years or so of surveillance. If you break that down further, there's probably 5-7% of the men who really have aggressive cancers, and should probably go straight to treatment, and this is the rare patient where we can really see clear evidence of progression in that short interval of surveillance. 

But then there's a very large pool of men who have absolutely no sign of cancer progression. The next biopsy is negative or we see one speck of 3-3 over and over again, and the PSA is stable or downgoing. And these are men that, frankly, don't need to be on active surveillance at all. They can be on a less active surveillance protocol. We could probably think about it more like watchful waiting frankly. And if we could validate a marker that could allow us to defer the next biopsy for four years, five years, that becomes highly actionable, highly valuable. It avoids side effects that avoids the risks, that avoids a lot of cost. I do think the markers will help us get to that point in tailoring surveillance. We're not quite there yet in 2018, but the studies that we're launching now I think will help us get to that point. 

Dr. Alicia Morgans: And I think, just to sort of tie everything together, as we improve our methods of diagnostics or diagnosis and we improve our understanding and ability to risk stratify patients, we continue to get the right patients onto surveillance and the right patients to active treatment. I am hopeful that the USPSTF guidelines will evolve with us, and will encourage us to actually screen men because, when they are screened, we'll know what to do with the information, and we'll be able to really match the treatment or no treatment with the patient and with his tumor. 

Dr. Matt Cooperberg: Well one would hope.

Dr. Alicia Morgans: Yes. 

Dr. Matt Cooperberg: They need to read the literature to get to that point, but yeah, there's still a comment in the task force statement about we can't distinguish indolent from aggressive cancer. It's just ignorant. Never mind the biomarkers. We can risk stratify with 80% accuracy based on stage grade, PSA, extent of biopsy involvement with an extremely well-validated risk stratification tool. 

The biomarkers that can help us do even better. But the notion that in 2018 with standard clinical information, we can't tell the rabbits from the turtles, it's just not true. Now, we haven't done a great job applying that information, and we're making progress as a specialty, but yeah, there's a number of steps. There's obviously changing clinical practice from the standpoint of urologists and radiation oncologists, changing the stance of the task force in the way they program their evidence review is a whole different conversation because the conclusions that they reach very much reflects the decisions that they make in their original methodologic program planning. 

As they establish what the rules of the road will be for what studies they will include and exclude, those decisions are going to drive their conclusions. There's a lot of us who think that they predict the endpoint based on which studies that use to include and choose to ignore, and hopefully they will continue to engage more and more with experts, with those who have actually been writing the literature in prostate cancer for many years, to come up with a better evidence-based guideline. 

Dr. Alicia Morgans: Absolutely. So, an evolutionary process, a process that will continue hopefully to change and grow over time and really meet the needs of our patients, because, at the end of the day, that's what we all wish for, I think. So, what are your closing thoughts, your overarching themes to leave the listeners with? 

Dr. Matt Cooperberg: Prostate cancer, like all cancers, is a complicated diagnosis and the devil is really always in the details. I think the problem is that label carries with it such a psychological impact with patients, that for many years, we've kind of had an uphill battle from the diagnosis because people hear cancer, and they think pancreatic cancer, they think months to live. 

I think the first challenge has been changing that conversation that just because we've used the word cancer, this is not that conversation. And then the second phase of that conversation is that within prostate cancer, we've got this extraordinary heterogeneity of risk ranging from life-threatening illness that really needs to be treated aggressively and with multimodal strategy to completely indolent biologically, virtually benign, that maybe we should not even label with cancer and really pinpointing for the patient in a way that he can understand where he lives on that spectrum or where he is on that spectrum continues to be a challenge. 

I think getting an individual man to the point where he can really understand where he is on the spectrum of risk remains a challenge. And those of us that really, eat, sleep, live and breathe nothing but prostate cancer from a clinical and research standpoint, even for many of us, it takes years to really get to the point where that conversation becomes routine, and where we really can guide that conversation reliably. 

I think in broader practice, it's a big challenge. I think the diagnosis is still carries with it this imperative to treat from a psychological standpoint, from a legal defensive standpoint, from a whole set of nonclinical factors that go into that decision. I think we're making a great deal of progress as a specialty in not only getting the message out within our treating community and then to the primary care community but to the patient community as well. 

And I think more and more men are kind of aware of active surveillance as an option. I think they're aware of prostate cancer as a not necessarily lethal diagnosis, and I think it's imperative that we all continue to work to really improve the messaging and communicate the messaging to our colleagues in primary care and to men in the broader community. 

Dr. Alicia Morgans: Absolutely. So really communicating to patients that you can live well if you have low risk disease on active surveillance if you have high-risk disease, that is a deadly disease, and we need to take care of it with whatever intervention is appropriate for that patient. These are important messages. And I do feel like we are aligning better as a field and hope that we can continue to work with the USPSTF and the data generation to do the best that we can for our patients. Thank you so much for your time.

Dr. Matt Cooperberg: A pleasure. Always a pleasure.