Elisabeth Heath: Thank you. Thanks for having me.
Tanya Dorff: We know that prostate cancer really behaves quite differently in different patients, and yet for a long time, our approaches have really been one-size-fits-all, but hopefully that's changing. So I wanted to hear your thoughts on some of the different prognostic tools that have been developed and are becoming more utilized, including the AI tool but also the sort of genomic tools.
Elisabeth Heath: Yeah. I think we can't go anywhere, especially here at ASCO 2026, without talking about AI. AI is sometimes biomarkers only. AI is sometimes huge large language model and now you have a new thing, whatever that thing is. Sometimes AI is really IT. So I think just as a general body of people, we're still learning on the go.
I know one of the times that it comes up a lot is when I see a patient with radiation oncology and they will do a referral and say, "What do you guys think about treatment intensification with ADT, whether it's very high-risk or high-risk and what are some tools there?" So we kind of get into those kinds of conversations. So I think right off the bat, it's not even in a metastatic setting. We're having conversations in a multidisciplinary setting.
Tanya Dorff: Well, that is where some of these tools have been developed and I think they're now just moving more towards advanced disease states. But we start out with those NCCN criteria that you alluded to, the high-risk, the very high-risk, the intermediate, favorable, unfavorable. And yet even within those, I guess we know that there can be differing cancer behavior.
So where do you tend to position something like Artera? We'll start with the AI tool. When are you ordering it? How do you find it useful in which kinds of decisions?
Elisabeth Heath: Yeah. You know, it's interesting. Our medical oncology group at Mayo Clinic, we tend to actually not order it and defer to our radiation oncology colleagues, but they don't also routinely order it if they're pretty certain of what they're thinking, meaning there's already not just one or two bad features, but they're kind of worried and concerned. They're definitely just leading to having that ARPI on board in addition to the ADT and the AI tool is helpful, but I don't think it's the ultimate driver of that final decision.
We spend a lot of time trying to go over what the results are. I think their website, for example, and just sort of how they derive the data is rock solid. I look at that and I'm confident myself, but it still comes down to that patient in front of you and you can't factor in all the comorbidities, and you and I treat patients with diabetes, other heart disease. And so then I think putting those in, it's just another tool in the toolkit, so I think it's helpful.
Now, I have to say at Mayo Clinic, we embrace this idea of AI to the nth degree. So it is utilized in every format possible, whether it's clinical care or just trying to highlight a better pathway and algorithm. So I think AI in general is here to stay.
Tanya Dorff: I will say that Artera, as well as some of the other diagnostic companies that we'll get to, do have very patient-friendly reports that I find make it easier to show the patient, "Well, here's this test we ordered and here's what it says about the potential benefits of intensified therapy," or maybe sometimes to reassure them. You have those very nervous patients and when you can show them that their risk of metastasis at five years is 3%, some of that information is really helpful in our discussions, I think, and we don't always have another source for that to really show a patient.
I also used to find some of the genomic classifiers helpful on the lower risk side in patients maybe who were on the fence about active surveillance or in that favorable, unfavorable, intermediate risk where not everything lines up and maybe I had some questions about whether this patient really needed longer treatment, ADT at all, ARPI.
So do you use any of the other genomic classifiers like Decipher or any of the others?
Elisabeth Heath: Normally when they come and see us in oncology and medical oncology, it's already been ordered typically by the urology folks. So they have what they typically order and what the radiation docs do and it's the same process. I think it's sort of, how does medical oncology weigh in?
I like the idea that we can start thinking about treatment de-intensification or not even just intensification, just treatment. I think patients like to see things like, "My score is this," much like they're looking at their PSA a lot. It's hard to interpret what's the blob on the bone scan or what does that SUV really mean on the PET scan, but when you're talking about, "Well, here's this chart." It's got some colors to differentiate if you're low or high. If you're in the intermediate, we all could just look at each other and go, "Well, and there we have it," and usually a laugh ensues and it's just another tool again. But I like the fact we have these things because it's just one more thing to help us guide what's the right next step.
Tanya Dorff: Well, I was really excited to see that Decipher was applied in some very large clinical trial data sets, like CHAARTED and STAMPEDE, and then we'll see here at ASCO ENZAMET also, because I think we are in the metastatic setting definitely looking for ways to differentiate who needs maybe triplet versus doublet and some of those questions. Any of those data that you find sort of practice-ready or you think it's still a little bit more in the research realm?
Elisabeth Heath: I think it's still a little bit more on the research realm. When I can't see it applied uniformly in a group, for example, it's hard to say, "This is our standard practice." I am looking forward to learning more at this meeting, see what those results are, but I think there, what a good problem to have. When you and I began many years ago, we didn't have too many options to talk about. Now we have options based on the actual patient's clinical status, some of the maybe not-so-nice biomarkers potentially. I think that's a wonderful use of the tools.
I think there's a lot of effort to try to figure out, even if it's the right answer, which is the better answer and as we continue to evolve, especially in the mCSPC state, I think that'll be even more of a topic. And so the conversations will change in the next probably year and how we talk about this with our patients.
Tanya Dorff: Yeah. I guess we're all anxiously awaiting some of the prospective biomarker integrated trials, like NRG-GU009, where we're actually using the tool to put patients into an intensification or de-intensification bucket and a randomization. So I think that would maybe solidify our ability to apply the tools.
I guess one question is, test versus PSMA PET, do you find that they're complementary, contradictory?
Elisabeth Heath: I think, again, they're just tools and it's up to the experts to put them in some context. Where I think we all could do a better job educating everyone is people glom these as just, "Well, I had some genetic test somewhere from someone." Or you ask, again, when we ask, "Oh, have you had genetic testing?" They don't really know what you're asking. So I think from a patient education level, it's also the awareness of that tool is different. People know a bone scan looks at bone things and, "Oh, there's a glowing number." As my patient says, "How much am I glowing on this one, Doc?" But when it comes to sort of these genetic sort of markers and tools, it gets really hard for them to understand what that means for them.
So I think it's also helpful to have these tools, as we were mentioning, that you can maybe refer them to the website that has a patient-friendly explanation because otherwise they're just looking at you saying, "Well, it totally makes sense if I'm right here talking to you, but then I have to go home and tell my family what you just said, and I'm not sure if I can do the same." So I think at some point, we all can do a little bit better, but we're also learning on the go too because in our group, when you're not routinely ordering but you're helping to interpret, that's also not as the same as if you're always ordering and then interpreting. So lots on both sides to improve.
Tanya Dorff: Yeah. I guess it's important for radiation oncologists, urologists, and medical oncologists to sort of put their heads together as they start to incorporate more complex information about a patient beyond just clinical parameters and maybe get a sense of how each other weighs those various pieces of information in making the final decision, otherwise it could get really confusing for patients if they hear sort of different interpretations or different recommendations once they have some of these more sophisticated tests.
Elisabeth Heath: Yeah. Well, there you go. Another UroToday discussion on a panel for all three groups and say, "How do you really interpret this information?" That would be quite interesting.
Tanya Dorff: Well, thank you so much for joining us today. It's been great talking with you.
Elisabeth Heath: Oh, absolutely. Anytime. Thank you.