Andrew Armstrong: Thanks, first of all, Neeraj, to have this conversation of how we translate clinical trial guidance into the clinic, because that's what we're going to do next week. To back up first, for working group two and three, I want to just talk a little bit about how that impacted practice already. That was 10 years ago, the last working group. In the past 10 years, because of that working group guidance, we have developed a number of therapeutic agents successfully in the clinic because we've understood that some patients, for example, will have a bone scan flare while they're responding, and we learned from that guidance not to remove our patients from therapy prematurely, to keep patients on despite new unconfirmed bone lesions, and we all know that. It's very common now to see a patient, for example, with metastatic prostate cancer. Their PSA drops from a hundred to two, they feel great, they're back golfing, yet they see new lesions on that first new bone scintigraphy scan. That's a flare phenomenon or pseudoprogression, and we learned the two-plus-two rule from the last working group to really require confirmation before we start transitioning a patient to a new therapy. And that was really responsible for the success of enzalutamide, apalutamide, darolutamide, and abiraterone now that we can use those in the clinic, and we've moved those drugs into earlier and earlier settings where we're seeing improved cure rates.
We're seeing much more impressive survival results. And so those guidance documents for protocols became the most cited in the past 10 years of any paper in prostate cancer because every clinical trial follows those guidelines. They really established uniform consensus criteria for eligibility, biomarkers, patient-reported outcomes, and imaging endpoints so that we can all do it on the same page. Working group four really builds on that. It develops those guidelines, moving them into earlier disease states, and we've also developed a new nomenclature, which we'll talk about. I know that's probably the most confusing part, but if you think about 20 years ago, there was a new nomenclature that we started to adapt that was also equally awkward. We started using castration-sensitive, castration-resistant. We didn't use to say that. That was, and it still is, a biologically correct term. Our patients who are progressing on ADT, androgen deprivation therapy, that is a form of medical castration, but I don't think you and I talk like that to our patients. We don't say those words. We say, "You're responding," or, "You're not responding to this therapy. We're going to now talk about the next therapy," and we're still going to be talking that way. A lot of this language is for publications, for research, for protocols, not necessarily the everyday language that you might use in a conversation with a patient.
Neeraj Agarwal: But they hear that. When they go online, they are looking for different resources for education. They see the word castration, which many of them, in my personal experience, do not appreciate.
Andrew Armstrong: Right. I agree with that, and that's why we've changed it.
Neeraj Agarwal: Yes.
Andrew Armstrong: So the working group four guidelines have jettisoned that term entirely. We've broken the guidance of nomenclature into three different categories. We have naive, sensitive, resistant, and that applies to any treatment, docetaxel, androgen deprivation therapy, a PARP inhibitor. We don't presuppose the outcome before it happens. That would indicate a naive setting. A sensitive setting would be somebody who's responding. The PSA is going down, they feel better, their tumors are shrinking. A resistant tumor would be something that's not responding. There's some manifestation of that disease that's getting worse. Usually, the PSA goes first, but sometimes it's the opposite. Sometimes the imaging gets worse and there's a discordant progression or the patient feels worse, so they're no longer clinically benefiting. We revolve around the mainstay, the front line of therapy, which are these androgen pathway modulators, and now we're using some of these AR pathway inhibitors, enzalutamide, apalutamide, darolutamide, and abiraterone, sometimes without ADT, without castration.
So what do you call that if the patient's testosterone is normal but they're on this monotherapy? Or let's say they're on an AR degrader or an AR RIPTAC or some of these newer agents that we're going to be developing in the next 10 years. What do we call that? And so we've come up with this broader term for androgen pathway modulator, which is a catchall phrase for anything that's impacting the AR pathway. Now, there's many other treatments out there, PARP inhibitors, molecularly precision therapies around PSMA. We've heard at this meeting a wealth of new antibody-drug conjugates or T-cell engagers, really exciting. We're going to need to redefine over the next 10 years the next-generation of cell surface targets, and maybe that will change how we talk about patients as well.
Neeraj Agarwal: So going back to our practice in the community, I heard the word androgen pathway moderation-sensitive prostate cancer, androgen pathway modulation naive prostate cancer.
Andrew Armstrong: Correct.
Neeraj Agarwal: And the old castration refractory prostate cancer terminology should preferably be replaced by androgen pathway modulated resistant prostate cancer. Is that correct?
Andrew Armstrong: That's correct, and then we layer on top of the APM term, whether it's metastatic or not, and then of course that's in the eye of the beholder. How you define that metastasis depends on your imaging. We know that some patients with conventional imaging have no metastatic disease, but the minute you do the PET scan, the PSMA PET, there's metastatic disease. So how you define the metastasis has to be indicated by the imaging modality, and so that's another layer. There's a third layer, which is the patient in front of you, their symptoms, what you see on the imaging, their pattern of spread. We know that patients with liver metastases have a worse prognosis than those with non-metastatic or node-only disease. And then a final layer would be the genotype, and we're learning so much more about DNA and RNA to classify patients, RNA expression, artificial intelligence classifiers, genotypes. There's a wealth of information, both germline and somatic that we can use. All of this is being done to try to improve the outcomes of our patients, to categorize the patients most likely to benefit from our therapies, to avoid those therapies that are not helping our patients, just causing toxicities, so that when we treat our patients with these therapies and they are no longer benefiting, we're stopping it, but if they're still benefiting, we're continuing. And we're just trying to get closer to a more perfect management style really using these guidelines, and that's impacting practice as well.
Neeraj Agarwal: Regarding the PSA progression criteria, we know a lot of our colleagues out there rely a lot on the PSA progression, and I found something very intriguing, that PSA progression criteria has changed in these guidelines, in this criteria, consensus criteria. So can you tell us more about how our colleagues out there should be looking at PSA progression moving forward?
Andrew Armstrong: Yeah, thanks for bringing that up. This was motivated by something that we've all been seeing called discordant progression, a patient who's otherwise doing well and you're celebrating that low PSA, but suddenly, there's radiographic progression or symptomatic progression. I bring up a case of neuroendocrine small-cell that I've seen in clinic where the PSA remains low, but there's clear full-blown disease, and we all know that. What we've learned from analysis of ARCHES or TITAN or some of the darolutamide trials or ENZAMET is that discordant progression is actually relatively common. A full quarter of patients on these AR pathway inhibitors, when they develop progression on imaging do so without even a change in their PSA, so zero change in their PSA, and the majority progress without this old definition being satisfied, this 25% rise or a rise of two. And so we're retaining that definition for standard of care if you're not on an AR pathway, but we want you to also report any change that's confirmed, but also to recognize that we can't rely just on PSA. So there needs to be regular imaging of our patients in order to capture that poor prognostic radiographic event. Now, most of the time, we celebrate a PSA decline.
Most of the time, a low PSA nadir, an undetectable PSA is great news for our patients, and most of the time, that means they're going to live longer and better, but we just have to be ... We can't just rest on our laurels there. We occasionally need to image our patients as these studies were done. Maybe we risk-adapt that, that we're not imaging every three months but perhaps every six to 12 months, so that we're not just losing sight of this discordant phenomenon.
Neeraj Agarwal: Any PSA progression in my view, based on the discussion, should be considered, should be taken seriously. It doesn't have to be two nanogram per mil and then adding 25% nadir, adding 25% increase to the nadir. Is that correct?
Andrew Armstrong: What I would say is this is not a guidance on when to stop treatment, but we do know that any change in PSA when you're on an AR pathway inhibitor might be a reason to get imaging because of that discordant phenomenon, but even in the absence of any PSA change, we should be getting imaging. Working group four is really trying to collect all this information so that in the end, we can be better doctors for our patients, so that we know the best trigger. It could be that working group five will be doing ctDNA monitoring over time, not just totally relying on PSA or imaging, and that would be nice to have a liquid biopsy test that gives patients reassurance that the therapy they're on is working, that they don't need to stop therapy. But we're not recommending stopping therapy because of PSA changes. We're just trying to harmonize definitions so that we're all speaking the same language.
Neeraj Agarwal: And lastly, it is quite common, even in my practice, to not do scans every three months, especially in metastatic hormone-sensitive prostate cancer, and now androgen pathway modulation sensitive prostate cancer. In castration-resistant or androgen pathway modulation-resistant prostate cancer, I do scans more often. So for our colleagues out there, how often they should be doing scans? I now know, these are just guidelines, consensus statement, but based on your own expertise and experience and being the lead author of this statement, what is your advice for them?
Andrew Armstrong: It's a great question. NCCN guidelines steer away from this entirely. They basically leave it to your judgment based on the patient's risk and your preferences. I would say imaging is needed. It's not needed, like you're saying, every three months in the hormone-sensitive setting. In a clinical trial, you may wish to do that because you're tracking a radiographic endpoint, but in the real world where a patient's doing very well, I tend to image at the completion of induction chemo hormonal therapy. For example, if I'm doing triplet therapy, I like imaging before and at the completion, so I'm re-baselining that patient. Usually, that's a celebratory scan because the patient's doing well, it's a remission scan, and you're using that going forward for all comparisons, and then every six months. The discordant progressions that we're publishing in these phase-three studies tend to happen in the first two years, and because it happens so commonly, 25 to 30% of patients experiencing that, I think more imaging in the first two years is reasonable. And then once you realize the patient's doing really well, I also, like you, back away from doing so much scans, and certainly we're not advocating for PSMA PETs every three months in every patient in the community practice.
For clinical trials where we're trying to document progression on imaging and trying to qualify an endpoint like PSMA PET progression, that may be different, and getting into who's paying for that standard of care and how to qualify that and the regulatory implication is probably a different discussion. In the APMR setting where risk is much higher, I think scanning with regular intervals is very reasonable. We often would scan every two to three months during the initial phases of treatment to document response and then back off on that a little bit every three to six months.
Neeraj Agarwal: What a wealth of information, and I think the best part I liked about the new terminology is finally, I will not have to use the word castration, a terminology, a term which I personally found to be not very liked by our patients.
Andrew Armstrong: I think many men would appreciate this change.
Neeraj Agarwal: Yes. So congratulations, Andy, and thank you for taking the time.
Andrew Armstrong: Thanks, Neeraj. Great conversation.