In this interview, Ian Thompson, MD, answers questions about the downside of diagnosing low-risk prostate cancer, the burdens of active surveillance and how to minimize them, and how his own approach to screening and biopsy have changed.
We want you to address some of the issues we face with low-risk prostate cancer. We have been challenged to a great degree with questions about whether we are screening too many patients with PSA, what to do about low-risk prostate cancer, and whether we should actually stop calling low-risk disease “cancer.” What do you think?
The challenges that we face with low-risk prostate cancer are not unique to prostate cancer. We published a paper in Lancet Oncology in May (2014; 15:e234-42) that showed very clearly that this is a problem across all cancers. The question of “Should we call it cancer?” is an issue not only for urologists, but also those who treat thyroid cancer, melanoma, lung cancer, breast cancer, and others.
It’s a little more of an issue for urologists because we have the PSA test and biopsies, and now we know unequivocally that the cancers we tend to find through screening-the ones that we increase the detection of-disproportionately tend to be such low risk that we know with a pretty high degree of certainty that many of them will never impact the patient’s life expectancy. That’s why the U.S. Preventive Services Task Force, for example, recommended against PSA testing. When these low-risk cancers are included in the group of men that you diagnose and if you treat them all, this unfortunately creates a greater harm than the benefit from detecting the high-risk cancers and treating them with multimodal therapy. That is why this is such a big issue that we really need to address.
Many urologists, myself included, see this as two separate forks in the road. Getting a PSA test is different from what we do with the information we obtain from it. What’s the downside of diagnosing these very low-risk cancers?
There is a huge downside, and I would contend that there are actually several forks in the road. The first fork in the road is, do you do a PSA? The AUA guidelines are very clear that if you are less than 40 years old, you probably don’t benefit. If you are 40 to 55 years old and low risk, the likelihood of benefit is quite low because in younger men, you are more likely to find low-grade, low-volume disease that rarely causes a problem. There are ways that you can actually predict who’s at greater likelihood of low-risk disease: men with lower PSAs, those whose PSA went up to a small degree, are very young and Caucasian, and whose rectal exam is normal. Those people preponderantly have low-risk disease.
The second fork in the road, I would contend, is also a problem. That is when your patient is found to have low-risk disease, thinks about it, and says, “I’m going to go on active surveillance.” Active surveillance is not a good thing. Patients choose it because they prefer the quality of life outcomes rather than those for radiotherapy or surgery, but active surveillance is a bad outcome. Patients repeatedly want to know their PSA results, even when they were just tested a week ago. They are still anxious. They have repetitive biopsies, sepsis, and so forth.
I’ve heard it said jokingly that PSA stands for “patient-stimulated anxiety,” and aside from anxiety, there are other risks of proceeding to transrectal ultrasound biopsy, including fluoroquinolone-resistant infection. You have a great calculator on your website that helps patients and physicians make these kinds of decisions. Tell us about that.
I use risk calculator all the time for that first fork in the road. The first fork occurs with the man who whose dad had prostate cancer, is Caucasian, young, and his PSA was previously 1.0 ng/dL. Now it’s 1.5 ng/dL. He asks if he should come in for a biopsy. The problem with those circumstances is that that man may have a 2% risk of a high-grade tumor, but he may have a 15% risk of a low-risk tumor. Out of a hundred men, there may be a two in a hundred chance that you might help that man, but there’s a 15 in a hundred chance-15 times greater likelihood-that you might find an indolent tumor and, by doing a biopsy, a 2% to 4% risk of sepsis.
When you present those data to him, he may say, “Maybe I don’t want a biopsy now. Maybe we should wait, repeat the PSA in 6 months or a year, and make a deferred decision.” You can also use that approach for the second fork in the road: treatment versus no treatment at the time of active surveillance decision making.
The online risk calculator, based on data from the Prostate Cancer Prevention Trial (PCPT), helps determine a man’s risk of low-grade and high-grade prostate cancer. Originally the results were just presented as percentages. The results are now also presented graphically with “Emojis,” which are pictures of faces that illustrate good outcomes, bad outcomes, and the outcomes in between.
The risk calculator is free and is constantly being improved. I use it in virtually every patient. It helps me help my patient make an informed decision.
Some European urologists are getting away from transrectal biopsies and going back to transperineal biopsies to avoid infectious complications. There have also been recent papers about the use of rectal swabs. How do you perform prostate biopsies in your clinic? Also, what are the other burdens of active surveillance?
This is one of the reasons we’re thinking about ways that you can unburden active surveillance. My standard approach, which is the approach used in the Canary Foundation’s prostate cancer active surveillance study, is PSAs quarterly, exams every 6 months, and biopsies every 2 years. I generally don’t do a first biopsy within 6 months or so, as long as I’ve had the pathology read and as long as it’s had an adequate number of cores. I may do it if the patient has something that doesn’t make sense, such as higher volume disease or possibly some component of Gleason 4 disease. Generally that’s what we do, but that’s a tremendous burden for the patient.
Regarding my approach to biopsies, I’m still doing a dozen cores transrectally. In patients who’ve had a prior biopsy, I also give them gentamicin. Now, I always ask myself if that is the right thing to do because it’s increasing pressure on the flora of the gut. But I saw an increasing number of patients who developed fevers and chills just on the fluoroquinolone, and I found that that substantially decreased. When a patient develops sepsis from a biopsy for active surveillance, that automatically takes away a huge amount of benefit that you’ve provided.
The other thing I am doing is reducing the frequency of biopsy in some patients. For example, in the patient with microfocal Gleason 3/3 disease that started in his early 60s, who’s now in his early 70s and has had two negative biopsies, I’m spreading it out more. We’re working with the Canary Foundation on developing a risk-assessment tool to be able to predict the person who’s at a low likelihood of coming off active surveillance, so you would be able to tell a patient there’s a 95% or 96% likelihood that your biopsy won’t find something that will take you off surveillance.
Does your follow-up of patients on active surveillance change if the patient has multiple biopsies because, as you know, performing multiple biopsies increases the risk of complications?
It’s important to point out that I have been taught as much about prostate cancer by our research as by our patients. I’ve had a number of patients who said they don’t want another biopsy and pointed out that their PSA and exam have been normal, and nothing has been found on three biopsies. They are some pretty smart folks. They say, “If that were a problem, you’d likely find it in this circumstance, and you haven’t, which means that I probably have a lower risk of high-grade disease than the guy on the street with a PSA of 1, so why are you repeating my biopsy?”
Do you feel there is any form of imaging that should be used now to help guide urologists, or are they all still experimental?
I think there are some circumstances where it may be helpful, for example, in the man whose PSA or rectal exam or biopsy are not in agreement with each other. One example is a man who went on a 5-alpha-reductase inhibitor, his PSA went down, it starts to go back up, and your repeat biopsy doesn’t find anything. Another example is the man whose PSA is rising and whose subsequent biopsy doesn’t find anything. Particularly if that man is younger, I worry, and I’ll order a 3T MRI with a coil. I’m not sure the coil is absolutely necessary, but there’s no question that a multiparametric MRI with a 3T yields much better images. If I don’t see anything, I don’t do anything.
My problem with this, however, is that the moment you do this, people will say, “If it worked for one guy, let’s apply it to everyone.” Then the problem comes in that the specificity of MRI in most series is about 60% to 80%, averaging around 70%. Once you take a technology that has a 70% specificity and apply it to a population whose prior probability of disease is very low, 30% of them will be false positive. Then what are you going to do? Are you going to do targeted biopsies in those men and take more cores? You are applying all of that additional burden and cost associated with repeating the biopsies on people who have very low risk to begin with. I am concerned that if we apply it in a fashion that is not individualized, there will be a net harm.
At what point do you add a saturation biopsy or a template biopsy? There are urologists doing as many as 70 cores. Is there ever a place for that?
I will do perhaps one or two or three a year. It’s very unusual. In general, I prefer to do the MR and if there is a lesion on the MR, if I’ve done an adequate number of biopsies elsewhere, I’ll just target the lesion. Sometimes they’ll be positive and there will be bulky high-grade disease, but there are also false positives as well.
Again, it really needs to be individualized. I think saturation biopsies tend to be a last-ditch resort. I don’t do it perineally, and I think you can do it adequately transrectally, except perhaps in the largest of glands. The difficulty from a technical standpoint is that if you push the needle way into the prostate and then fire it, it often doesn’t get a throw. We push the needle way in, pull back about a centimeter, and then hit a throw, which provides the momentum to obtain an adequate biopsy. In general, if the MR is negative, the likelihood that there’s a high-grade tumor there is really low. The negative predictive value is 90% to 95%.
Are there biomarkers that you’re excited about that may help us prevent biopsy in patients who don’t need it or help determine patients who do need it?
There are some tissue-based tests-the Oncotype DX test from Genomic Health and the Prolaris test from Myriad-that can be used to predict tumor grade or recurrence. PCA3 and ProPSA are also available, and percent-free PSA, fascinatingly is pretty inexpensive and is one of the most powerful biomarkers to indicate the presence/absence of cancer. T2ERG may be something that’s useful, and there’s a host of other analyses, such as urinary-based methylation studies, that are coming into play.
A more recent, fascinating one that Elizabeth Platz and Angelo De Marzo looked at in the PCPT sample is something very cheap: presence of inflammation in the benign part of the prostate. In patients with low PSA levels, they’ve shown it has an odds ratio of four, meaning inflammation elsewhere in the prostate increases the likelihood fourfold that there’s high-grade cancer present.
Currently, I think you can use the PCA3, and you can use one of the tissue-based tests. You might also use the presence of significant inflammation in biopsy cores to make a recommendation. But it needs to be done appropriately. When you use predictive markers in a very low-risk population, your likelihood of a false-positive result is far greater than a true positive. It really needs to be limited to the patient in whom something-his PSA, exam, his risk factors, or biopsy-doesn’t feel right.
There are actually statistical ways that can be developed. We’re working with Rice University, for example, to develop a chip-based approach that looks at a broad range of markers and then adds additional markers based upon what’s called the likelihood ratio. You can use artificial intelligence to put several risk factors together and then make a recommendation on when to use markers. You don’t want to use additional markers in people who have very low-risk disease. The other way of looking at additional markers is that they are probably a bad idea in people who de novo have very high-risk disease as well, as they could give you a false sense of security that you don’t need to treat them.
In this country, there are obvious concerns about medical litigation and what you can and can’t tell patients. What do you tell patients when they reach age 75, and when do you feel comfortable turning these patients loose?
Oftentimes the patient tells you what they want to do, and that’s my preference. You can’t be 100%. A good example is an 80-year-old patient who was followed by his primary care physician, and had very low PSA levels over a long period of time. When he turned 80, everyone was in agreement that he probably no longer needed PSAs; 2 years later, his PSA was in the 400 range. So you can never be 100%, and the patient needs to understand that.
But for the man whose PSA is stable and biopsies don’t show much of anything, the utility of continuing the PSA long-term is very, very low. If you think about it, most people who die of prostate cancer have high-grade disease at the time of the original biopsy. The paradox is that, when you sit down for the first time with a man with low-grade disease, you can tell him he has a lower likelihood of dying from prostate cancer than the man in the general population who doesn’t have it. Once they begin to understand that, they’ll tell you what they want to do.
In your practice, with all the studies you’ve done, how has your approach to PSA screening changed?
I think the thing that I’m better at doing now is identifying the man who’s most likely to benefit. Consider a 45- or 50-year-old man who walks in with some voiding symptoms, and I start him on tamsulosin. Historically, we would tell him to get his PSA checked on the way out.
Now, the conversation isn’t too long but goes something like this: There’s a blood test and there are pros and cons of it. The advantage is that if you have prostate cancer, there’s a greater likelihood we’ll find it. The disadvantage is that we may find tumors that are inconsequential, and treatment of those inconsequential tumors can lead to side effects. I ask them if they’ve ever had a PSA test, if they would like to have one, or if they’d like to learn more.
That short conversation is now part of my practice. I make sure that the patient understands he is not just being offered a blood test, but he is embarking on a journey. Some people may say they absolutely want a screening test because their neighbor died of prostate cancer. Others might say, “You mean, if I get a test that finds a tumor that might not be consequential, and the treatment for that inconsequential cancer might render me impotent? I’m not interested.” And then there are people in between, and you can provide them more information to help them make a decision.
I find that what patients are looking for is direction. I don’t find many people who don’t have some bias, but they generally ask me as the expert, “What should I do?”
I would respectfully disagree. I find the number of people who turn everything over to me is really low. Look at the newly married person, for example. He and his spouse are concerned about longevity, and sexual function is really important, so the potential detection of an inconsequential cancer that can lead to sexual dysfunction is a real issue. In the man who is no longer is sexually active, it’s completely different.
You can pick up in the conversation with the patient and often the spouse how much information they want. You can help them make the decision, but I’m uncomfortable sending that low-risk man for a PSA without at least telling him that there are some risks and benefits. I find that patients may be initially uncomfortable with me helping them make a decision but ultimately truly appreciate it.
In a man who’s 60, are you even bringing up the discussion of PSA screening?
Absolutely. The AUA guidelines state that men who are 55 to 70 should have informed decision making. I cannot ignore the evidence of the fall in mortality-and some of that benefit may be due to improved treatment-there’s incontrovertible evidence that screening makes a difference. To not tell that patient those results is unjust. Outside that age group-the younger man or the older man-there are some other risks that they need to understand.
Do you have anything else that you would like to add?
First, I suggest again the use of the online risk calculator and tell patients about it. Second, when a patient is referred to you who’s had a low PSA before and has a spike in their PSA, repeat it. PSA has a high degree of variability. Can you imagine going to an internist who finds your blood pressure is a little high and immediately puts you on an antihypertensive? Repeat the PSA, and you will look like a genius.UT
Subscribe to Urology Times to get monthly news from the leading news source for urologists.