Keywords: prostate cancer, biopsy, PSA, biomarkers
How to cite: Loeb, Stacy. “When to Biopsy: The Role for PSA and Novel Biomarkers” Grand Rounds in Urology. January 21, 2016. Aug 2020. https://grandroundsinurology.com/biopsy-role-for-psa-novel-biomarkers.
When to Biopsy: The Role for PSA and Novel Biomarkers
As you know, the traditional approach to determining the need for prostate biopsy was just based on an absolute PSA threshold, that is, applying one-size-fits-all approach with one PSA value for everyone. So, that’s how it was approved by the FDA back in 1994, and that’s how it was used in the randomized studies of screening for prostate cancer, which were also designed in the early 1990s based on using just this one cut-off value for everyone to determine biopsy.
The problem with this that we’ve since learned so much more about PSA and the use of this, is that it’s not like a pregnancy test. I wish it was. My patients wish it was. Everyone loves things that are black and white, where it’s either you see the pink line or you don’t. But unfortunately, PSA really reflects the continuous risk of prostate cancer, and no better evidence on this than from the prostate cancer prevention trial where biopsies were done irrespective of PSA levels.
You can see that even at low PSA levels, they were below the traditional cut-off, the risk of prostate cancer increases continuously with PSA really across the whole spectrum. The same is true for high-grade prostate cancer.
The other issue, of course, is that PSA is affected by so many other factors, so it just is not the ideal test, and you know about these; age, race, having an impact of course, prostate volume and enlargement, androgen levels, obesity being linked to lower PSA levels. The type of assay you use–so we ran the same blood sample using different assay standardizations, and the results differed in the same sample by up to 20%. So, you can imagine over time you could get a pseudo-accelerating or deceleration just by using different labs for the PSA.
Some medications can have a small impact on it. Genetics; we found that some men have higher PSAs just based on their genotype, so maybe the screening of the future will involve genotyping patients. Benign prostate conditions, obviously urinary tract infections, manipulation, et cetera.
Clearly there’s limited specificity for prostate cancer and all kinds of other things that could have an impact on this. But, this is a problem because when we get an abnormal PSA, that triggers invasive follow-up testing, and we know the increasing risk of complications from prostate biopsy, not to mention other downstream effects like over-diagnosis leading to over-treatment.
So, what we need are biopsy triggers with greater specificity. We can do such a better job of this now. I agree with all the previous speakers. This is not the time to just–we’re coming upon the end zone; we’re not going to turn back around and run across the field the other direction.
We need to just keep moving forward with all the progress that we’ve made in our understanding. There’s better ways to use the PSA test, adjusting for some of these confounders. There’s new PSA-based markers. There’s new markers that are not related to PSA. We have better imaging and we have multi-variable tools. So, let’s keep moving forward.
First and foremost, just very simple stuff. Thinking about the impact of age on the PSA levels, and this nothing new, this was proposed by Dr. Osterling back in the early 1990s, this concept of using higher PSA thresholds for older men to take this age factor into consideration.
This is not dead; in fact, even the correct AUA guidelines suggest considering a higher threshold of 10 for men over 70, which I find particularly helpful, I must say. Of the entire AUA recommendations document, that might be one of my favorite ones, because it really helps to dissuade elderly men who are less likely from benefitting from wanting to have a biopsy that made do more harm than good.
Another way of adjusting PSA to make it more specific is PSA density. So, reducing the confounding for prostate volume, and if it’s above 0.15 this does indicate a greater risk of prostate cancer and aggressive diseases. Now, this has been somewhat limited, at least for initial biopsy, because traditionally we always did these calculations with the TRUS volume, and who is going to do a TRUS before the time of biopsy?
But things may change; now a lot more men are having MRI, so if you do have a prostate volume measurement available it’s possible to do this even before the initial biopsy. Actually, you can even make a very gross estimate of the volume from DRE. It’s not very good, but you can tell if the prostate is small or big, and make a very limited judgement on this.
PSA velocity; so this has gotten a lot of controversy, but these trends matter. We can’t just ignore the medical data that’s in front of us, so if the patient hands you a list of their last five PSA values, no doctor is going to tell them to throw that in the garbage. So, we do want to have a way to interpret these trends. Actually, the truth is that most men with BPH, it is not just skyrocketing over night.
So, if you look at some of the randomized studies of BPH, for example, and the placebo arm, and the control groups of men with moderate to severe BPH symptoms, typically the PSA velocity is around 0.15 per year. So, it’s not 0.4 or 0.75 per year; it’s much lower. Those types of rates of change would be much more consistent with prostate cancer.
Of course, if it spikes so much more than three in a year, then that suggests that it’s more likely to be prostatitis. But this is still not perfect. Not all men do come in with a list of their last six PSA values. As Vickers has shown, if you try to use only two tests that are separated by four years or more, in all of his studies it did not work. If you don’t have enough PSA data, then you’re not going to be able to use PSA velocity effectively.
Free PSA; so, I’m sure everyone here is very familiar with how PSA is circulating in two different forms, free and complex. This is dating back to the early 1990s, but this is still around. In fact, as you saw in the earlier slides, the 2015 NCCN guidelines still mention free PSA as a potential reflex test prior to an initial or a repeat biopsy.
This leads on to the next subject, which are some of these PSA-based marker tests. So, this is a schematic explaining what all these different forms of PSA are. So, actually what we now know is that there’s lots of different isoforms of PSA, and pro-PSA is more specific for prostate cancer. This forms the basis of the prostate health index.
This isn’t really a new marker; really what it is is just a mathematical formula that’s weighing the different concentrations of these types of PSA components. It looks super complicated with the square root and everything else, but what it’s saying is that if there’s more pro and there’s more PSA and the free is lower, that’s bad. So, if you think about the equation, that’s basically all that’s happening.
So, at any given PSA concentration, what’s happening with the amount that’s free and the amount that’s pro? That is what goes into this formula. This has been demonstrated to be more specific than PSA and large multi-institutional prospective U.S. studies, European studies; it’s been validated in Asia.
It predicts more aggressive disease at biopsy, at prostatectomy, and serial measurement of PHI, predict progression on active surveillance. We showed that in the Johns Hopkins Surveillance Program. This is out there; it’s FDA-approved since 2012, and it’s included in the NCCN guidelines for a reflex option for initial and repeat prostate biopsy.
This is the basic idea of PHI; you’re still getting all your separate values. For people who are not familiar with what the interpretation is of the test, you still see the PSA, in this case 4.4, with 21% free PSA, but it’s that pro component here that’s really high. So, the PHI is 97 on this example, which is a super-high value indicating a very high risk of detection, and especially of high-grade disease on a biopsy. So, this is a situation where knowing that pro value gives you more information than just from the free and the total.
Let’s go to our first audience response question. What actually are the three components of the Prostate Health Index? Can you identify which one on this list is not part of the formula?
Okay, good. Wow, everyone’s really listening, that’s great. Yes, intact PSA is not. It’s total, free, and pro; intact is part of the 4K score, which is our next topic.
So, the next PSA base marker test is the 4K score. This is a similar concept to the Prostate Health Index where we’re using these different isoforms of PSA, these different Kallikrein markers, but now we’ve got four of them instead of three. So, hence the name 4K; 4 Kallikrein’s.
Actually, the other difference between PHI and 4K that not everyone appreciates, I think, is that 4K is already taking into consideration age, rectal exam, and previous biopsy history in the algorithm. So, the formula is not just the PSA base markers, but also these other factors that are influencing prostate cancer risk. The company does the proprietary formula and what you get back is the risk of finding high-grade prostate cancer on biopsy.
Again, this also has improved specificity for overall and clinically-significant prostate cancer. It’s been shown many times in European populations. More recently it was validated in a large prospective U.S. study showing that it outperformed the PCPT risk calculator for high-grade disease. Also it predicts aggressive disease at prostatectomy.
You saw earlier some of the data about prediction of future metastatic disease. This did not use the FDA pathway, but it’s CLIA test, it’s commercially available, and also is included in the 2015 NCCN guidelines.
PCA3; this is the next category of markers. This is a urine marker now instead of blood tests, which is all of what we’ve discussed so far. This was actually first discovered to be over-expressed in prostate cancer tissue, and then it was later found that this could actually be measure in the urine. Many different studies of it showing that it predicts the risk of prostate cancer on repeat biopsy better than total PSA levels.
It was approved in 2012 by the FDA, same year as the Prostate Health Index, but this one only for men who already had a previous negative biopsy. Both the Prostate Health Index and the 4K score can be used prior to any biopsy. This one was specifically improved for prior to repeat biopsy because it didn’t perform as well in the initial biopsy setting.
My main concerns about it, though, are actually not so much that, but the fact that it’s not really clearly linked to more aggressive cancer. So, many conflicting data on whether it actually predicts aggressive disease. As I see it, the problem is finding significant cancer, not just finding any cancer, which we already do a pretty good job of. Head-to-head studies have actually shown that the Prostate Health Index outperforms PCA3 for detecting aggressive cancer, so I think that’s where we need to concentrate our focus.
But, it’s not over yet for the urine tests. We heard a nice lecture earlier this morning with all the work at University of Michigan of potentially combining some of these urinary tests together. So, combined with other markers like temp reserve, or perhaps this SChLAP1 in the future may be a multi-plex urinary panel. It could perform better than PCA3 does on its own.
Audience question 2; all of the following are considered options for both initial and repeat biopsy in the 2015 NCCN guidelines except…?
Okay, great. So, PCA3 is only listed in the NCCN guidelines as an option before repeat biopsy, whereas 4K, Prostate Health Index, and free PSA are listed in both settings.
Now, another option is available, and I’m not sure why this wasn’t initially listed in the NCCN guidelines for repeat biopsy, but this is a tissue test. Interestingly, most of the tissue tests that we discussed at these meetings are focused on tumor tissue, but this one is the only one looking at benign tissue.
So, what we’re using ConfirmMDx to do is determine does the benign tissue have changes of hypermethylation that suggest an occult cancer? This tissue test is checking benign tissue to look for these methylation changes, suggesting a field effect around a cancer that the need just didn’t hit the first time around.
You saw this report earlier in Dr. Crawford’s update. What’s nice about ConfirmMDx, as you’ll see, is that you actually get a picture. So, just like doing an MRI gives you not only the idea of who has greater risk, but also gives you a map of where to look for the tumor.
This one gives you a map because you see, here are the methylation changes, we’re at the left base and the left lateral base. So, if you were going to do a repeat biopsy, clearly those were the locations that you would want to focus on. It’s nice that it gives risk information and localization information.
Which segues nicely into MRI. This isn’t really the subject of my lecture, but it wouldn’t be really complete without at least mentioning the huge expansion and improvements in imaging technology, standardization of scoring with the PI-RAD system. If you do have a PI-RAD 3, 4, and especially 5 lesion, you’re at a much higher risk of prostate cancer detection. We do a lot of MRI-targeted biopsies at NYU, and I think this is the future. It has been shown to be cost effective in both Europe and in a U.S. setting, at least before repeat biopsy.
Just to summarize with all these things, all of these new markers are definitely big steps forward, but we have not found the golden bullet. We need to consider multiple factors together to make the best decisions for our patients.
Everyone has a different amount of risk based on their history and their risk factors, and there are some risk calculators available that can help conglomerate these variables, whether it’s the PCPT risk calculator online, giving it more information on the risks of finding prostate cancer on biopsy, or this app from the Rotterdam Group, where you can enter some of the values. It’s under Rotterdam Risk app in the app store.
You can use some of these if you want, manual assistance, in trying to put together some of these risk factors to give the patient a more individualized estimate of the risks that you’ll find cancer or aggressive disease.
In conclusion, we do now have multiple variations on PSA with greater specificity, including PSA density, PSA velocity, free PSA. There are several new PSA-based markers, both the Prostate Health Index and the 4K score predicting higher risk of high-grade disease on biopsy, both part of the 2015 NCCN guidelines. PCA3 is another FDA-approved test. Some concern about whether it predicts aggressive diseases, but it is an option. Maybe combined with other urine tests in the future.
ConfirmMDx tissue tests, looking at negative biopsy to predict a cancer that was missed, and an increasing body of evidence supporting MRI as having a role in biopsy decision-making and in targeting of the biopsy.
Finally, a one-size-fits-all approach like we saw at the beginning with total PSA is totally out, and multi-variable approach is in. So, we’re in the era of personalized medicine; we know how to do a much better job of this now. Let’s keep moving this forward.