PSMA PET Appropriate Use Criteria "Presentation" - Michael Morris

February 9, 2024

At the 2024 UCSF-UCLA PSMA Conference, Michael Morris highlights the importance of PSMA PET in prostate cancer management, acknowledging the regulatory framework established by pivotal trials that led to FDA approval for PSMA PET tracers. Dr. Morris discusses the challenges of interpreting PSMA PET results, the need for consensus in its application, particularly in biochemical recurrence and radioligand therapy, and urges further research to explore its full potential while cautioning against delaying established treatments for emerging strategies.

UCLAUCSF video logo stroke


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

Read the Full Video Transcript

Michael Morris: And Johannes, Jeremie, Chuck, Tom, wherever you are, thank you so much for this conference. It's really so nice. But also, thanks to everybody who's flown from far and wide to come together. This is such a nice, intimate, academic experience. We really appreciate it.

All right, we're going to talk about PSMA PET appropriate use criteria, and I think we can save ourselves a little bit of time by not repeating what everyone has said about certain slides.

So first of all, a good place to start in terms of talking about consensus criteria is just the regulatory framework. And as we know, we have the three trials that led to the staging indication, and whether you have three trials that led to the detection of relapse disease indication. And in essence, the FDA approval language is very similar for each of the three tracers, and that is pick your generic tracer is indicated for patients with suspected prostate cancer metastasis who are potentially curable by surgery or radiation. And also, for patients with suspected prostate cancer recurrence based on elevated PSA levels.

I think we have to be very grateful for the huge ambiguity of what any of these things actually mean. There's no definition of suspicion for metastatic disease. There's no definition, really, of what relapse means, except that at some point in the disease, the PSA is rising. So actually, the agency has given us a lot of flexibility to practice, as doctors and as investigators, to study these various places where PSMA PET actually has regulatory endorsement.

Tom went over these trials very well, and I'm not going to repeat what he said regarding sensitivity and specificity, but what I would just add to what he said is not only are their populations somewhat different, but the endpoints are as well. The endpoint for LIGHTHOUSE was somewhat more rigorous from a histologic standpoint than the other two.

I guess the only thing I would also add to what Tom said about these three trials is that any study in which the comparator is microscopy, any imaging study is just not going to look that great. You're going to have a lot of false negatives until, at the point at which PET is essentially equivalent to a histologic diagnosis, with a step-sectioned specimen looked at under a microscope. You will never have sensitivity equivalent to that when that's the gold standard. And so, it's not surprising at all that the sensitivity is modest.

But I also think that as we were just talking about using PSMA PET for targeted therapy, I think somebody asked this question, remember that if you do not see disease, it doesn't mean there's no disease there. You're just not seeing it. So I wouldn't take this to imply that if you're doing targeted therapy, that you're not missing something because an area is not seen. On the other hand, most of us are faced with positive predictive value as our clinical dilemma. We don't know if the patient actually has disease or not. All we know is that the patient is relapsed or that the patient has a new diagnosis of high-risk disease. All we have is the scan. And if the scan is positive, despite our excellent talk on the false positives earlier this morning, it's more likely that the disease is there than not, for a positive lesion.

And then Tom really talked very eloquently about PROPSMA. The only thing I would add to this, to what he said is, A: I totally agree, this is an outstanding study because it looks at the strategy of imaging what you do first. And it not only speaks, as Tom pointed out, to the increased accuracy of PSMA PET, but the efficiency of getting to the right answer quickly, with one study, as opposed to the other pathway of conventional imaging and then having to sort of do makeup studies in order to really sort out what's going on, that that too has clinical relevance in terms of both financial economy, economy for the patient in terms of fewer imaging sessions, and the patient's own personal economy.

So I think all of those speak to the message of PROPSMA, which is that it is more both accurate and, in many respects, better for the patient as well, and maybe society, to do PSMA imaging upfront.

But as we discussed, if you do image for staging purposes with PSMA PET, you do increase the likelihood of detecting synchronous metastatic CSPC. And as a result, then, I think you get a lot of ambiguity from consensus criteria for that dilemma. So all of the... I'm not saying that this is a comprehensive list of all consensus criteria developed for PSMA PET, but these are some of the major ones organized chronologically. And I think everyone agrees low-risk disease should not be imaged. But in terms of intermediate disease, risk disease, some consensus criteria split these, some do not split these, but in the high risk or very high risk, where you might think that every group would have consensus on doing PSMA PET upfront, that's actually not the case. And the reason for that is the clinical concern that if you picked up synchronous metastatic disease that's only detectable by PET, that without knowledge of what best practice is, that PET patient would be denied definitive local therapy, especially with surgery, because now you're in a situation where you have a new set of facts and you don't really know what to do with them.

And I think that that's why, for example, AUA and as well as ASCO, had this issue that maybe we should be getting a standard set of scans so that we know what at least we should be doing had we not had the PSMA PET because with the PSMA PET, we don't know what to do. And that is a real dilemma.

Now on the other hand, SNMMI, EAU, and NCCN have said it's fine to do, from a consensus standpoint, PSMA PET as your upfront study as the only study, and use that for the clinical decision-making that you need to.

For the biochemically relapsed population, again, the performance characteristics are excellent for disease detection in this clinical state, in terms of going back to the original data. But what I think that there are significant clinical implications here, which everyone has touched on, first of all, the biochemically relapsed population, some of those patients are going to get upstaged into outright metastatic disease. The low-volume patients by standard imaging may get upstaged into high-volume disease. And for those who are saying, "Would you ever get a PSMA bone scan?", this is why you would get it because we don't have a treatment paradigm based on multiple phase three really well-conducted studies that have led to the demonstration of increased clinical benefit distinguishing between high and low-risk disease. We only have that for bone scans. So that's why we order bone scans still, in this context, because our clinical guidance hasn't made its way into the PSMA PET world without those data.

The non-metastatic CRPC patients get risks, have a risk of being upstaged, and just outright metastatic CRPC. I personally don't see that as a risk. It's the same treatment paradigm one way or the other.

And then the risk in the metastatic CRPC situation is that we use PSMA PET, or if we were to misuse it, to declare radiographic progression. We use radiographic progression as a key, not only in clinical trials endpoint for registration studies, but we also, we've convinced the rest of the community effort to prostate cancer working group two and three to not change therapy until you see radiographic evidence of progressive disease. But, now, if you declare rPFS at essentially the same time point as biochemical relapse for metastatic CRPC, you're doing exactly what we moved away from, which is at the first indication of a change in the disease status of the patient, i.e., early relapse, you're discontinuing treatment. And that's really risky in terms of having good, life-prolonging therapies that we know when we stopped therapy early, we did the patient a disservice. It was a real training period to get the community to wait for radiographic changes, and now we're going to essentially move those radiographic changes two months earlier.

So there's risk there. And I think that that's why for the consensus criteria for the advanced patient in that metastatic CRPC space, you see a focus on using PSMA PET for lutetium candidacy, not so much for assessing disease outcome, especially for non-radioligand therapy.

For BCR, you see almost uniform consensus that clinically speaking, this is an appropriate place. Some criteria have carve-outs that are just for patients who are candidates for salvage RT, or just for higher-risk patients. But generally speaking, I don't think anybody disagrees that BCR is the ideal place for PSMA PET for this indication.

Metastatic CSPC, again, for those who commented on it, there was this concern of up-voluming the disease, and for non-metastatic CRPC. There's a fair amount of heterogeneity in terms of the appropriate use of metastasis-directed therapy, but not a lot of heterogeneity in terms of the use of PSMA PET to pick up relapse, which is an interesting thing. But I think it's the metastatic CRPC space where, really, PSMA PET has a unique ability to determine treatment candidacy.

As we discussed before, there is not consensus necessarily on whether all tracers are equivalent or not. SNMMI consensus criteria said yes, EAU said no consensus. Although if you look at those criteria, like there are three different questions on this, and everyone answered each of the three different questions three different ways. So, depending on how the question was phrased, the answer was either yes or no. But I take that as a lack of consensus.

For selection criteria for candidacy, most said VISION is fine, but use your head, as well.

For on-treatment response monitoring, this is interesting. Not a lot of consensus on this, although I think as I'll show you in a few minutes, I think it's a very provocative use, especially when you're going to do something with that monitoring, in terms of dose adaptation.

And then, finally, in terms of on-treatment assessments for progressive disease, no consensus, and there was a general sense that this was not ready for prime time.

I think I just want to talk a little bit about PSMA and its role for radioligand therapy, since there is consensus that it is useful for candidacy and treatment selection for patients. We have a number of different nomograms based on retrospective data, a phase two study, and a phase three study, in which all of them indicate that PSMA avidity increase is associated with clinical outcome with lutetium PSMA-directed therapy. And in particular, the VISION, which is the only phase three trial that's prospectively done with a survival outcome, PSMA PET is the most influential, in terms of predicting patient outcomes.

So I think that there are some important roles there for selecting treatment, but it doesn't mean that this is the only treatment available to choose from. And that's what I think that tomorrow, when we talk about the TheraP trial, chemotherapy versus RLT and how to make that decision, is going to come up.

Louise presented at ESMO a really nice slide presentation on Enza-p, and how PSMA PET can be used to adaptively dose radioligand therapy. I think that this is something that we should definitely, as a field, adopt as a research question, and really build into as many trials as we possibly can, because I do have a sense that a one-size-fits-all serves everybody poorly. And if we can figure out a more patient-individualized treatment plan based on PSMA PET, that would be a good thing, and I invite everyone to ask Louise about this, post this presentation, because I don't have time to go through it, but she's sitting right here.

So for oligometastatic disease, for non-metastatic CSPC, and metastatic CRPC, the application, of course, is what we just discussed, in terms of metastasis-directed therapy for each of these. But I would just caution, because I did hear somebody say, this will allow us to delay giving systemic therapy. Just remember, there has been no trial where giving systemic therapy earlier hasn't prolonged more life. So before we go, "Oh yeah, that's great, I delayed giving systemic therapy that prolongs life," just make sure that, actually, that that's the case, because just delaying life-prolonging therapy in and of itself isn't necessarily clinically valuable. It's a terrific strategy, but we need to recognize these are small, small studies, much smaller than the studies that we have used to prove the utility of early systemic therapy.

So just to wrap things up, in terms of appropriate use, I think it's incontrovertible that we should do no imaging for low-risk disease. And that given how you approach the clinical conundrum of the staging PSMA of what to do with that information, you can use it either alone or as an adjunct to other imaging modalities to try to sort through what the staging would be with and without the PSMA PET to understand what the risks that you face for a given patient, in terms of making clinical decisions based on those differences. Everyone agrees that the BCR patient should get a PSMA PET. Treatment selection for radioligand therapy, everyone agrees on.

Everything else is open to debate, discussion, and certainly to more data collection, analysis, and discussion. Thanks very much.