Podcast Transcript: Use of Genomics for Active Surveillance

[Dr. Aditya Bagrodia]
There's nothing long-winded about it. It's fantastic and it's thorough. I do think that probably to keep it somewhat digestible, we'll focus on or maybe our favorable intermediate risk and below cancers, and then we'll also focus on people with at least a 10 to 15-year life expectancy. Then, of course, there's going to be an extended spectrum, right? Your 82-year-old is at three heart attacks and a stroke and comes in with a grade group 5. Okay, fine, we'll watch you, but all right, fantastic. I love it. You started out with family history, germline testing. The numbers I have are 12% metastatic, 4% or 5% localized. I do think dialing in on that history, as you mentioned, is critical. One of the things that I actually find a bit confusing to patients sometimes is comparing and contrasting genomic testing and germline testing. Does that ever come up in your chats with patients?

[Dr. Ashley Ross]
Yes, it comes up all the time. What I explain to them is, well, and actually I should say more for that, it comes up with providers all the time too. Right now, the NCCN, when it looks at things that predict how the cancer is going to behave, it looks at prognostic tests. It gives a level-one certification of evidence for genomic tests, specifically, Decipher is the only one in that category. Then, it talks a little bit about Artera AI, which is a newer test that hasn't yet been validated outside of radiation patients, but that looks at the digital pathology. Then, it has a level 4 evidence, meaning there's just one study not corroborated about genetics, but that genetics that they're talking about, those studies are looking at not only things like BRCA, but we're seeing the emergence of polygenic risk signatures that can tell you what's your SNP signature that's going to tell you the individual beyond your family history, if what's going to happen in terms of developing a cancer. Regardless, for your actual question, providers are getting confused in there too, but for the patient, what I tell them is your genetics is what you inherited from mom and dad. It drove their predisposition to developing this cancer, and more so, in some cases, the cancer itself can be driven by that genetic abnormality. Genomics is better said as a test looking at what your cell is doing. What's it doing in terms of what genes are turned on? What's driving it? I don't go this deep with the patients. They usually accept that and we go to the next step, but sometimes when patients are really confused and they say, just explain it to me even more simpler, I tell them that, look, your body DNA is like the Library of Congress. You're not allowed to check out any books from the Library of Congress. It has all the information in the Library of Congress and that's your DNA, but the DNA is all the knowledge. Then, what your cell is doing when it wants to do something, it goes to the Library of Congress and it starts Xeroxing pages, like the instructions of what to do, and it takes that away. Those Xerox copies are your RNA. If I want to know what the cell's doing, I can look at the RNA. What are they Xeroxing from the Library of Congress? If the patient still looks confused and they're like, well, I still don't get it, and I'm feeling a little bit punchy, I'll say, hey, have you seen that movie Seven? They say, oh yes, that's a great thriller. I'm like, okay, how did they catch the killer? How did they know it was him? They were like, oh, they looked and saw, what books did he check out of the library? I was like, great. The books they checked out of the library in that case was the RNA, the whole library is the DNA. We want to know, are your cells killers? Did they check out the wrong books from the library, or are they not? I got to feel pretty punchy to have the discussion derail to that point. Usually, I'm just telling them that genomics is looking at what the cancer, your cancer specifically, is doing right now. Or what its biological potential is. Not like, what did mom and dad give you that then caused this seed of cancer to develop?

[Dr. Aditya Bagrodia]
I like that, and I might have to start using the Library and the Seven analogies. Those totally resonate.

[Dr. Ashley Ross]
I love it.

[Dr. Aditya Bagrodia]
Germline testing, genomic testing, which we'll talk about, there's the MRI, there's the cores, there's a core involvement, the pattern for the variants, PSA density, which I'm sure is also something that you're factoring in. There's a ton of information that we have. One practical question, especially for the folks coming in from the outside with their biopsy in the community, not MRI, and let's say you get an MRI, nothing overly offensive, does surveillance for you start at that time of initial biopsy, or do you like confirmatory biopsies? Or is it case by case?

[Dr. Ashley Ross]
It's a great distinction you're making because even in the literature, for those that go through the literature, sometimes you're looking at literature where they're talking about a "surveillance series," but it's actually studying the period of time where you're qualifying that man for surveillance. A lot of times, just to your point, I'm telling the gentleman, like, we want to figure out, are you actually qualified for surveillance?

Let's say the guy comes in and has anything but very low-risk prostate cancer. Very low-risk prostate cancer, I know the AOA is getting away from that definition a little bit, but the NCCN definition of very low-risk prostate cancer, which just, if there's trainees and residents on, or just as a refresher, was originally developed to find out what would be the pathological correlates on a sextant biopsy that would equal like half a centimeter or so of low-grade disease. To reach those criteria, you had to have, of a sextant, less than 3 cores involved, less than 50% of a core, all Gleason grade group 1. We know a lot about those guys. We have more than 20 plus years of information on those guys. All that information says it's safe to follow them and that information came even before MRI was used in practice, right? Those guys also, from the beginning of your talk, and I think you were setting us up for this, I almost never diagnosed that in my practice because I use my MRIs and stuff. An MRI, again, for the group, is not to, it is to get the best biopsy you can get when you do a biopsy, but more so as a public health issue is to eliminate doing a lot of unnecessary biopsies, right?

Those guys almost never exist, but let's say everybody else, let's take our guy with like three or more cores of Gleason grade group 1, or our guys with 3, 4, 7 disease, grade group 2, I'm telling them, first, I need to qualify you for surveillance. The evidence says that we can watch this. If they had not gotten an MRI, I usually do an MRI about six months after their initial biopsy, sometimes earlier if I really think something is wrong, but I want to survey them, so sometimes three months. In less than three months, I will find a lot of artifacts, so six months is usually my goal. Sometimes three months is early. Then I do a fusion biopsy to qualify them for surveillance.

Let's make it easy on us and pretend the guy came in and they've already had the MRI, they've already had the biopsy, and now we have their biopsy results, and then we're going to talk about qualifying them for surveillance. The reality is that surveillance should be offered to basically everybody with favorable intermediate-risk disease and less. It definitely should be offered, but as I said before, what the patient wants to know is where are they going to be on these curves. Are they going to be on the curve that's flat that they never need treatment their whole life? The ProtecT study, other series have suggested that for low-risk disease, that might be up to 50% of the patients. Of favorable intermediate-risk disease, that might be up to 25% of the patients, depending on age and other factors. Are they going to be on the disease curve where it's going to take them a while to progress, but we're looking at like three years or more of surveillance before we have to think about progression, and that might be a win for the patient of good sexual function, et cetera, and progression will be slow enough that we can treat them to cure. Or are they going to be on the curve where within one year or a couple of years, half of them or more are going to have progression to something we really don't want, with 4+3=7 disease, T3b disease, disease that is going to be less curable?

In this regard, the person who comes in, and we're trying to figure out what's what, and I should tell you, we don't absolutely know, because, in the surveillance trials that have gone on, they've had faults in two ways. One way, we were just understanding surveillance since the '90s till now in the PSA era, and we were scared about following cancers. Even if they got a little bit worse, people were coming off surveillance. That was one problem, so we really don't know, you know, when to do what. Then the other problem was, in other series, they were allowed to progress on surveillance in a way that at least, this side of the pond or in the US, I would find sort of unacceptable, like lifelong ADT or something from a guy who was curable is not a win for me. Particularly as we extend surveillance into my guys, there's no age cutoff for me. Even if you're 40 or 50, you can be on surveillance in my practice, but as I extend it out, I want to know, well, what's our goal? I break down the risk factors into risk factors dealing with disease biology and risk factors that deal with disease extent. The disease extent risk factors, in my mind, I give them like one point or yellow flags. Those are things like PSA density, which is a measure of disease volume, PI-RADS-5, which is a measure of disease volume like I mentioned before, 1.5 centimeters or EPE, number of cores involved. Those are all one point. Then, the two-point things are things that have to do with your disease biology because these are things that give you metastatic potential, like incurability potential. Those things are, we mentioned intraductal cancers, cribriform can be more aggressive, if the guy was a known BRCA2 mutated guy, but then importantly, then one of the topics today is their genomics. Specifically, Decipher is the most validated, like I mentioned, NCCN gives it level-one evidence backing its ability to be prognostic for metastatic progression. Most of us around the US are using Decipher as our genomic test of choice. If that's high risk, it can be bad news bears. I'll just give one example, and then we'll go, we can take a deeper dive into it.

In the MUSIC Registry, which was Michigan Consortium of Urologists Academic and Community, there was a nice paper by Randy Vince and colleagues, and they looked at people that were going on surveillance that had genomic testing with Decipher. Decipher, again, it's a molecular test looking at from the biopsy, or you can do it on radical too, but in this case, we're talking about biopsies, where you're looking at the RNAs, and you're running the whole RNA profile of the genome, but you're looking at this 22 gene signature that has been validated as the best prognostic signature for metastatic progression and it's been valid in other endpoints as well. Now, if you take favorable intermediate-risk patients and low-risk patients, in that MUSIC Registry cohort and in general cohorts, it's somewhere about 15% of the patients, it's ranged between 10% and 20%, but 15% of the patients are going to have genomic high risk. We look at the genomic high risk in that cohort, by one year, half of them had come off surveillance, by three years, 75% had come off surveillance. That was about, I think, two to three-fold higher than the people who had low genomic risk or intermediate genomic risk. Then you could say, well, maybe the patients just got scared. They saw this high-risk genomics and they got scared, and that's why they came off surveillance, but then actually, the telling figure is that for the people who came off surveillance and went to treatment if their genomic scores were high, their chance of biochemical progression, remember everyone we started out on surveillance, their chance of biochemical progression was threefold greater than the guys that had low or intermediate genomic risk. Even more scary, although the numbers get small, at their three-year follow-up, almost 40% of them had biochemical recurrence and needed subsequent treatment. That means that those guys need, now went from a one-punch cure to one, two-punch, and maybe they're still not cured.

Then, if you look at the other series that took people with genomics that were otherwise Gleason grade group 1 or 2, so like you're a low and favorable intermediate risk, had genomics, and then you took them to radical. They were people that maybe a little bit of a right-shifted population because you were going to take them to radical. This includes our Northwestern series that's in preparation right now, but if you look at series that came out of like collaborations in UCSF and Hopkins or ones from Yale, if your genomics is high risk and you're otherwise Gleason grade group 1 or 2, the chance that you have 4+3=7 disease in that prostate or seminal vesicle involvement is 40%, as opposed to 10% if you're genomically low risk. 40% chance you're following something that none of us would want to follow in someone with a 10-year survival. I think that's what's pushing that group into this sort of red flag area.

One last thing I'll say, there was a nice study at the AUA. It was a kind of a SEER registry that is finally linked into genomics. Jim Hughes' group and others that made the first publication on the SEER registries, like 2000 patients that had genomics with Decipher. They looked at all the different things that would predict that you would take a patient off surveillance or take them to treatment if they were otherwise low or favorable intermediate risk, and those things were what we talked about. Disease volume was high. PSA was a little bit higher. Genomics itself, Decipher was high. The only thing that actually predicted that 4, 3, 7, or T3b pathology was the high genomics. Actually, the volume of percent pattern 3, even the small amounts of pattern 4 were not as predictive.

Again, it's long-winded, but essentially when we're qualifying for surveillance, I look at my risk count and I look at risks that have to do with volume and risks that have to do with disease risk: the cellular risk. If you have a score of more than 3, then I tell you, look, surveillance at your own risk, we're going to definitely do a confirmatory biopsy at one year, because this gets even more nuanced. We don't know how often we got to do confirmatory biopsies and evaluations, but if you have these higher risks and you want to do surveillance, we got to do the biopsy at one year, even if you already were qualified on an MRI. You have to tell the people that, look, there might be a 50% chance you're going to have some progression that's meaningful, and it's unlikely that you're going to be on surveillance for five years plus. We have to think about that 10% to 15% population, I have to really think about being pound foolish, like Pennywise pound foolish. The flip side, genomics, I used to be really scared with Klotz's data, I used to get really scared about following favorable intermediate-risk disease. If genomics look good, if your volume looks good, I've got those guys on surveillance for the last five years, and I've been really pleased to see that I'm not getting burned. We might have to break that into pieces, Aditya.