Optimal Imaging Standards for Metastatic Castration-Sensitive Prostate Cancer - Michael Morris

May 23, 2024

Michael Morris discusses optimal imaging standards for metastatic hormone-sensitive prostate cancer. He explains that conventional imaging methods like PSA tracking are insufficient for monitoring disease progression in the era of advanced androgen receptor signaling inhibitors (ARSIs). Dr. Morris emphasizes the need for regular imaging, recommending scans every six to twelve months to detect radiographic progression. He highlights the potential of PSMA PET imaging but cautions that its use is complex and not yet fully endorsed for routine follow-up. Dr. Morris advises individualizing imaging schedules based on patient risk factors and treatment responses to optimize care.


Michael Morris, MD, Prostate Cancer Section Head, GU Oncology, Memorial Sloan Kettering Cancer Center, New York, NY

Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, MA

Read the Full Video Transcript

Alicia Morgans: Hi, I'm so excited to be here today with Professor Michael Morris, who's joining me from Memorial Sloan Kettering, and also joined me at APCCC 2024 in Lugano, Switzerland, where he talked about how we think about imaging in the context of metastatic prostate cancer. Thank you so much for talking with me today, Michael.

Michael Morris: Well, thank you for having me, Alicia. It's a real pleasure to be on UroToday with you again.

Alicia Morgans: Wonderful. Well, let's get into your talk, and then I'll ask some questions after you finish.

Michael Morris: Today I'm going to talk about optimal imaging standards for metastatic hormone-sensitive prostate cancer, and what we would recommend in daily practice, and why. These are my disclosures. Conventional imaging in metastatic castration-sensitive prostate cancer used to be a relatively straightforward and easy-to-describe set of standards. We would just follow the PSA because, historically speaking, when patients were responding, they would have a declining PSA, and when they were progressing, they would have a rising PSA. And the PSA essentially was the leading indicator of what was going on with the patient. Whereas the disease burden, reflected by scans, was a lagging indicator. And symptoms, especially in the setting of progression, lagged even behind the scans. And so, it was pretty straightforward in those older days; you would just follow the PSA to know when to get your next image.

Imaging has always been a difficult thing in prostate cancer because of the bone tropism of prostate cancer. And we know that, historically speaking, our conventional imaging standards like bone scans and CT scans have not been particularly good indicators, especially of response. Bone scans, for about up to 50% of patients on active therapy, will flare as a measure of increased metabolic activity. In the context of a regressing tumor in bone, it would mislead us as clinicians, so we would think that patients were progressing when, in actuality, they were responding. And it was for this reason that we really dismissed response in our lexicon and focused on time to event endpoints in clinical trials. As well as when to change therapies, we focused on radiographic progression-free survival. We had standardized this definition for clinical trials in Prostate Cancer Working Group 2 and 3, and although that wasn't necessarily intended for standard of care, it did set the stage for pretty much knowing what progression looked like. An increase of two or more lesions was considered to be significant evidence of progressive disease, and we wouldn't act on the first post-treatment scan, in case flare phenomena were present.

That's not really the case now, in terms of patterns of response and progression. And I'm going to use as an example, Andrew Armstrong's analysis of the ARCHES trial. The ARCHES, to just remind the audience, was a randomized study of folks with metastatic castration-sensitive disease, and they were treated either with ADT and a placebo, or ADT and enzalutamide. And if you look at the data of these patients, and mind you, these patients were scanned every three months per Working Group 3, quite a number of patients progressed, on the basis of scans, without an antecedent rise in PSA. So if you look at those who progressed with no rise in PSA from nadir, that was just under 40% of patients, and just under 70% of patients progressed radiographically before meeting the progression definition biochemically.

So if you just followed PSA and didn't follow scans on a regular schedule of just imaging patients every three months, you would miss a lot of progressive events. Or you would at least diagnose them much later. And is that clinically significant? Well, if you look at the patients who progressed by scans without meeting progression criteria by PSA, it is clinically significant. Those patients do have, and these are the patients that we're talking about, the patients with a stable PSA, where their scans were progressing. These patients had a median overall survival of just about two years; you can see it's around 26 months, so that's really not very good. And so you do want to switch these patients' therapy so that they get onto more active therapy and hopefully have some life-prolonging therapy added to their regimen, on the basis of their radiographic progression.

So the investigators from ARCHES, in this presentation at Geno-ASCO in 2022, I think correctly concluded that regular imaging is recommended to detect radiographic progression among those patients treated with potent ARSIs like enzalutamide and others, because serial PSA monitoring alone wasn't sufficient to detect progression in these patients. And I would follow that same recommendation. But of course, these patients were scanned quite frequently, every three months. And I think that we don't know how many important cases of progression by scans alone would be lost if we made a more liberal schedule of every six months, or every nine months, or even yearly. So we're missing that piece in terms of thinking about an evidence-based decision-making algorithm, as to how frequently to image patients who have castration-sensitive disease. Every three months may be for clinical trials, but in standard practice, could we get away with every six or nine months? Probably so, but we just don't have the data.

And I would say that NCCN does recommend routine imaging in castration-sensitive patients for those treated with an ARPI, recommending imaging every six to 12 months. But again, the data is sparse in terms of knowing what that sweet spot is.

But what about PSMA, PSMA PET imaging as an alternative, or tomorrow's version of routine imaging for castration-sensitive disease? I mean, PSMA PET does actually image the tumor directly based on PSMA expression, and that's really the rub for these patients. So PSMA does have variable expression in castration-sensitive prostate cancer, and it's not too dissimilar from what we have seen historically with bone scintigraphy. We have two very nice early small reports, one from Tom Hope and the other from Louise Emmett, showing the variable PSMA expression of folks treated with ADT who have metastatic CSPC.

In Tom's series, he showed PSMA flare after about a month of ADT. In Louise’s series, she actually showed PSMA downregulation in those patients. So it's kind of confounding to know what to expect in someone who's responding to therapy, who has metastatic CSPC. Some patients will flare, others might get better without necessarily having anything but PSMA expression to account for that. It's not necessarily the volume of disease.

There is a nice report looking quite systematically at 25 patients who had metastatic CSPC, and these patients were followed by PSMA PET imaging at baseline, and at three to four weeks, by PSA of course, and as well, by FDG PET imaging at baseline and three or four weeks. And you can see here, in terms of the bone metastases, there were 406 bone metastases, about 26% of them did flare as a result of treatment with ADT. And that flare was really real. The median SUV change was 50%. And those patients also, you can see them here, are having declining PSAs, the Delta-PSA is the X-axis, the Delta-SUV max is the Y-axis. These are the patients who have declining PSAs but rising SUVs. And these patients generally, when you look at their FDG scans, don't have that much metabolically active disease. And so these patients may represent a group of patients who actually are going to do quite well on ADT, yet they'll have early changes that are unfavorable on their PSMA scan.

Now, how about progression? Those are the confounding issues for response. While in progression, these patients who are on an ARPI, remember that prostate cancer, just like other solid tumors, has this phenomenon of lineage plasticity, which represents a lineage switch away from AR-driven disease. And that occurs in about 15% of metastatic CRPC. And these are patients who are coming out of CSPC, which is what we're talking about, and into CRPC. About 15% of these patients will have this heterogeneity of non-AR-driven disease, if not partially then completely. And these patients will not produce PSMA.

I'm showing you here a patient with neuroendocrine prostate cancer, who's imaged with both a PSMA PET scan and a PET scan we're developing at MSKCC, which images something called Delta-like ligand 3. This is a small cell directed tracer. DLL3 is a small cell target. In fact, I've shown a small cell lung cancer patient here with DLL3 positive disease. With small cell lung cancer, those patients have really quite homogeneous DLL3 expression. With neuroendocrine prostate cancer, it's somewhat less homogeneous, but you can see how PSMA really doesn't show very well this patient's neuroendocrine emerging disease, through an ARPI. The better imaging modality for him is a scan that's most appropriate for the small cell lung cancer patients.

So what I'm showing you here is that PSMA, as a response indicator, can be deceiving. PSMA PET as a progression indicator can also be deceiving, in terms of what's going on. Of course, you do have a CT component here, to show you increasing disease, but to characterize that disease, PSMA might not be your best tracer in a lot of patients.

And it's for these complexities that PSMA PET, right now, is not recommended by most medical societies and consensus groups for treatment follow-up and evaluation for response or progression. It's a big maybe; we have a lot more to learn about PSMA PET in this context. It is appropriate to perhaps assess disease volume in metastatic CSPC if you're considering metastasis-directed therapy; molecular imaging does have a role in guiding the decision-making around metastasis-directed therapy. But in terms of follow-up for patients to detect either response or progression, there's a lot more that we need to do in order to learn how to best use PSMA PET as a tool for evaluating responders versus progressors in a serial manner after patients start ADT and an ARPI, which, after all, is the standard of care for many of these patients, if not most of them.

So I'm just going to conclude by saying serial imaging should be performed in the contemporary era of ARPIs. Monitoring by PSA alone really isn't sufficient. I can't tell you what the best imaging algorithm is based on data, but I think that it's safe to say that following the NCCN guidelines of six to 12 months is the least that you could do and will detect disease that's perhaps progressing by scans and not by PSA.

And PSMA PET is not a panacea for this problem of not having ideal imaging for the patient with metastatic CSPC; there's a lot to learn. We need to learn that on a context-specific basis. So distinguishing the CSPC patient from the CRPC patient and distinguishing patients who have CSPC who may be receiving an ARPI versus docetaxel, and, in the future, other therapies as well. So there's a lot to work on in terms of using new imaging or newer imaging modalities as a replacement for CT and bone scintigraphy in this context.

And, thank you very much.

Alicia Morgans: Thank you, Michael. That was really great. And actually, some points that I don't think we really emphasize and talk about enough, and I'm really glad to see that the NCCN is endorsing this routine screening, or imaging over time, for patients with metastatic hormone-sensitive disease. So I wonder, it's a bit of a broad window, six to 12 months. As you think about it in the clinic, which patients do you really err on the side of six months, or maybe even less? Which patients do you think maybe you can go up to a year in? How are you thinking through this, and how does the patient's clinical picture help you make some of that decision?

Michael Morris: That's a really great question, and it really points to how we approach patients individually and make decisions based on the risks that they may be progressing quickly, and might be progressing by scans versus PSA.

Let's take the patient that we're nervous about, a younger patient with an unfavorable mutational profile. That might be somebody who is BRCA, or has other HRD alterations. Someone who has a p53 mutation or a PTEN alteration or an RB mutation. So those are the patients that, characteristically, we are concerned about a poorly durable response to ADT and ARPI, for whom they may be at risk for early conversion to a neuroendocrine phenotype of one type or another. And so those are the patients that I really do get more aggressive with in terms of routine imaging.

How about the other side? Someone who, let's say, has had a long course towards getting their metastatic disease. So they may have metachronous disease, with a slowly evolving course, over a long period of time, favorable PSA kinetics. They have a nice response to initial ADT and an ARSI. They nadir down to less than 0.2, let's say. They may have a SOX2 alteration, also lending itself to a durable response to AR maneuvers. And those patients, you might say, "This patient has a favorable clinical presentation, and a favorable genomic profile, and he's nadired favorably, as well," so those are the patients you might err on the side of a longer interval between scans.

And then there's, of course, between those extremes, all of the gray zone that we struggle with all the time.

Alicia Morgans: Yeah, no, but thank you for talking that through, because I think that we're making these decisions in our clinic all the time, but it's nice to kind of just acknowledge that we're doing it, and we're kind of restratifying these patients, and that that's probably the right thing to do, so we're not over- or under-scanning some of these folks.

So here's another question that comes up in my clinic a fair amount. This is not exactly what you talked about, but definitely adjacent. Let's imagine the patient who is low volume, metastatic hormone-sensitive disease, low volume metastatic, based on PSMA PET. So you initially thought it was a high-volume patient, but now you see it as sort of an oligometastatic picture.

Imagine that we're treating this patient with whatever systemic therapy you think is right, probably ADT and ARPI. Probably not chemo, but if you want. And then you could have your radiation to the prostate, pelvis, and lymph nodes, and then some MDT to some additional sites. If this happens, how do you follow that patient? Do you use a PSMA PET to follow that patient? Do you use standard imaging?

And I have to say, I don't know that there's a perfect answer here, but I am faced with this, and different practitioners do all kinds of things. And we're going to take your word as gold, so you tell us all what to do, Michael.

Michael Morris: Sure. And we also have to work within the framework of the US healthcare system, and we can't just order what we want. We have to order things, generally speaking, within the label of those tests.

So let's take that patient. So first of all, you included in that, metastasis-directed therapy. So for me, I want to do that with the guidance of molecular imaging. Why do I want to do that? Because we do have data from the ORIOLE trial, for example, that when you add... Or the EMPIRE-1 study as well, that when you add metastasis-directed therapy, although EMPIRE-1 was focused more on the salvage radiation context. And for that matter, your question brings us out of context from ORIOLE, as well.

But when you're doing metastasis-directed therapy, I use molecular imaging to make sure that I'm actually irradiating all of the metastases. And also, I will probably hold off, if the image shows that the patient actually does not have oligometastatic disease, and would mean that I really am not, I'm just leaving a lot of disease behind, unirradiated, if I were only irradiating those lesions that had a corresponding area of sclerosis on the CT. So I've gotten a baseline PSMA PET, in that scenario. I have that as a comparator.

Alicia Morgans: Yes.

Michael Morris: Now, the next PSMA PET that I can order and which NCCN has guidelines around, is at the next progression event, in order to establish its distribution of disease. So I can't really order serial PSMA PETs, and as I just talked about, I'm not sure I would be able to trust those, either.

But the next time I order a PSMA PET will be in the context of some evidence of progression. And there is a study, that's the ARTO trial, which looks at metastasis-directed therapy for the oligoprogressive patient who has metastatic CRPC. And again, it's a slightly different context than the one that you just described, but we rarely get patients for whom the decision-making exactly fits the studies that we have.

So that ARTO trial also did support the use of molecular imaging, as a way of guiding metastasis-directed therapy for the oligoprogressive patient. It's all the scans in between those two events, though, for which I don't really have endorsement for using PSMA from a reimbursement standpoint. And as I just discussed, from a data standpoint. And so for those interim scans, I'm generally ordering just a CT and a bone scan.

Alicia Morgans: I completely agree with you, but it's so interesting that in so many practices, PSA can be undetectable, but everybody wants a PSMA PET, because we did some radiation, and we've got a PSMA PET to start. And it's just very interesting to me, because the PSA is undetectable, I'm not expecting that it's going to be a valuable scan given this situation. In addition, it is not an indication for that scan. So, anyway—

Michael Morris: I think, though, that there's a really good research question embedded in your case, there. And that is, you put this same patient, let's say, on all of those therapies: an ARPI, you irradiated the primary, plus or minus the pelvis, and the metastases. Okay? And the next question is, what is the resistance lesion, to all of those interventions, that's going to be tomorrow's source of lethal clones? So I do think that there's a research role for a PSMA PET there, to offer consolidative therapy to the resistant disease that looks like it's still active after you've applied all that therapy. But that is not a question that we know the answer to, and so, that shouldn't be part of standard of care.

Alicia Morgans: Great. Very good. These are the gray questions that come out of the APCCC, and they're in our clinics every day. And as we wrap up now, thank you for talking me through that thought example. What is your recommendation to clinicians, summarize your talk in a sentence or two?

Michael Morris: Sure. The pearl here is, one, for your metastatic CSPC patients who are on ADT and an ARPI, per today's standards, you should not strictly rely on PSA to declare the patients not responding. And assess, serially, using scans, the patients for early progression by scans without a PSA rise.

Number two, individualize your decision-making. Patients come in with pre-treatment features like their molecular profile, their Gleason score, their distribution of disease. And then, as you're following them, their post-treatment or on-treatment features, like their nadir PSA, what their plateau is, how long it's taken to get even some rise in PSA. Use those pre-treatment features, and those post-treatment features, to individualize your decision-making so that you're not over-imaging your lower-risk patients, or under-appreciating progressive disease in your higher-risk patients.

Alicia Morgans: Fantastic. And a final thing I would add is that there is still so much work to be done in understanding how to best utilize these imaging strategies. And at the end of the day, when we work with our teams and we work with our patients, we're going to hopefully get to the right answers, and answer some of these areas of gray data as we continue to try to do the best that we can. So thank you so much for talking us through this. Always such a pleasure to talk to you, and to learn from your expertise.

Michael Morris: The pleasure is mine, Alicia. It's always good to be on UroToday, and to be talking with you.

Alicia Morgans: Thank you.