Zachary Klaassen: Hi. My name is Zach Klaassen, a urologic oncologist at the Georgia Cancer Center in Augusta, Georgia. I'm delighted to be joined on UroToday ASCO 2025 in Chicago by Professor Nick James at the Royal Marsden Oncologist. Nick, thanks so much for joining us on UroToday.

Nicholas James: Pleasure, as always. Yeah.

Zachary Klaassen: We've got some exciting data you presented at ASCO looking at ArteraAI and AI model and looking at it in M0, high-risk prostate cancer patients from the STAMPEDE platform and really dicing out who we can treat and intensify who we can't.

Nicholas James: Yeah, absolutely.

Zachary Klaassen: So maybe just tell us a little bit about the ArteraAI prostate test. And where it's been, it's been a lot of info in the last couple of years.

Nicholas James: Yeah, so we've worked, I mean, our groups worked with Felix and so on for a long time. So sadly missed-- tragically missed actually. I mean, lovely guy. And as probably the listeners know, the test was trained on patients in early stage prostate cancer trials around the question of who needs hormone therapy on top of radiotherapy. So really left hand end of the risk spectrum. And the test will identify those men who benefit from having hormone therapy added in that setting. And you get a numerical score out told MMAI.

And so what we've done is-- so within STAMPEDE, they're all high-risk patients. So they're all on hormone therapy. So the test is therefore high on all of them because they all benefit from it. And within the abiraterone parts of the trial. We had two, ABI (abiraterone) only, and ABI plus ENZA (Enzalutamide) in combination. We did a pooled analysis of the M0 patients asking the question does adding ABI with or without ENZA improve metastasis free survival as the primary outcome and overall survival as a secondary outcome? And the reason we looked at MFS in that setting was that when we'd done the same analysis with docetaxel, it was negative. We didn't see benefit. And so we thought, well, it may be that ABI and ENZA won't work as well in this setting either.

Zachary Klaassen: Sure.

Nicholas James: But it turned out they worked really well. So what we ended up with was the MFS hazard ratio of 0.5. We halved the risk of metastasis or death, and the overall survival hazard ratio is about 0.6. So 40% improvement. But the patients although they're high risk a lot of them are not dying of prostate cancer. And therefore you're overtreating them, which is different to the metastatic setting where everybody lives a bit longer. So everybody's benefiting irrespective of where they sit on the spectrum of risk and from the disease itself.

So we hypothesized that because the MMAI tool is telling you something about sensitivity to hormone therapy that it may be able to tell us something about who was and was not benefiting from the addition of abiraterone. And so we ran two sets of tests with it. The first is it's prognostic. So within the high-risk M0 we've got two groups. We've got locally advanced with no nodes. So high PSA, high Gleason, high T stage, at least two of those three things. Or they're node positive. And the node positives do worse than the node negatives.

So within that applying-- we got a range of scores out from the MMAI all high, really high compared with the parent setting where the tool came from. And we did various modeling to see how we could split them up. So we split up them into median and split them into quartiles. You pick your split.

Zachary Klaassen: Yeah.

Nicholas James: And essentially the mathematical best fit for distinguishing good actors from bad actors was the upper quartile versus the lowest three quartiles.

Zachary Klaassen: OK.

Nicholas James: And so if you apply that for each of these groups, the node negatives it splits them into two groups with really quite distinct prognoses and the same for the node positives. But then the next thing which is really much more interesting, is we then split those curves by plus or minus ABI. And so for both of those curves, there's a benefit from adding ABI. But when we split the curves again by plus the top quartile versus the rest, what you can see is that almost all the ABI benefit sits in the top quartile. The bottom three quartiles, A do relatively well with the standard therapy, and B get very little delta from adding ABI.

So basically you can run this test and you can say, well, actually all of the benefit sits with just 25% of the patients. Because basically, if you're not going to die of prostate cancer, you don't benefit. And we find them. So it's a very pleasing result.

Zachary Klaassen: Yeah, I think, if you look at the Artera history going from the localized, we saw adding long-term versus short-term ADT.

Nicholas James: Yeah.

Zachary Klaassen: Last year looking at RP and whether we need to treat and intensify. Now we're getting into the high risk M0. And it's a perfect marriage of a very good test with very good long-term follow-up with STAMPEDE.

Nicholas James: It's fantastic.

Zachary Klaassen: When we look at the clinical implication of essentially saying, if you're biomarker positive, you need intensification or you're negative and you don't. How would you explain that to patients? What's the implication for that?

Nicholas James: Well, to a degree it's a black box. We don't know what's happening. We don't actually know what it's doing. But it's reasonable to, I mean, I haven't had to try and explain it to patients yet, so that's a good question actually. Good.

Zachary Klaassen: That's why we have these interviews so we can hypothesize a bit.

Nicholas James: Yeah, yeah, good thought experiment.

Zachary Klaassen: Yeah.

Nicholas James: So generally true of cancer drugs that we give the same drug to everybody knowing that only a subset of patients are benefiting. And yeah, what we always are saying to patients, if we had a test that told us which ones were going to get the drug, we'd use it. And so we've published other stuff from STAMPEDE using, for example, RNA expression profiling that shows that we can do something very similar to this with docetaxel sensitivity, which actually is separately interesting. So we know there's intrinsic differences in biology that drive whether you respond or not. We just don't know how to identify them.

So what the ArteraAI test is doing is it's identifying something about the biology that predicts sensitivity to hormone therapies in general, and very specifically abiraterone in this example. And exactly what the biological features is picking we don't know. We don't actually need to know for the purposes of applying the test. But for me, the other thing that's so interesting about it is we have the physical samples. We can go back and interrogate those now try and figure out what's the biology that's driving this because that's potentially new druggable targets.

So it's interesting on two separate levels. One is a practical exercise. We can-- looks like we've got a pretty good way of identifying who's benefiting and who is not benefiting. And particularly for the node negative patients, if you're node negative and in the lowest quartile, you've almost got no risk of dying from prostate cancer with just a standard treatment. So for 75% of the patients will be able to give the really positive news.

Zachary Klaassen: Yeah.

Nicholas James: Yeah, they don't need two treatments and they're going to get a good outcome. And for the ones who've got the highest quartile test, even though they've got a relatively good prognosis, still we can make it better.

Zachary Klaassen: Yeah, now it's another step towards precision medicine, which is awesome. Very exciting data. Thank you, as always, for spending time with us on UroToday Dr. James.

Nicholas James: It's a pleasure.

Zachary Klaassen: Thank you.

Nicholas James: Yeah.