Read the Full Video Transcript

Oliver Sartor: Hi, I am Dr. Oliver Sartor with UroToday, and I have a special guest here, Ravi Madan, who has published many things of interest, but in particular, talking about PSMA PET. Now, let me introduce Dr. Madan. First of all, he's a senior clinician at the National Cancer Institute and the genitourinary branch. Very well known among GU circles. First of all, welcome, Ravi. Glad to have you here.

Ravi Madan: Thanks, Oliver. I feel like we always get to have these conversations at meetings, so it's nice to do one of these as well.

Oliver Sartor: Well, really a pleasure to have you. In the JCO, you published a provocative piece on PSMA and how seeing more could potentially be detrimental to care. Let me ask you, first of all, to sort of define why you put this manuscript together, what were the motivating factors, and then broadly your conclusions. Then I'm going to come back and query some of those points. Take it away.

Ravi Madan: Sure. I think my co-authors and I, Rick Lee from Mass General, Edwin Posadas from Cedars-Sinai, and Evan Yu from up at Fred Hutch, and then my NCI colleagues here, Dr. Karzai and Pete Choyke, we kind of have been talking about this and it's come up as we've experienced referrals from the community or second opinion-type cases, and it became striking how frequently in the community they were bypassing CAT scans and bone scans in patients with metastatic castration-resistant prostate cancer and using PSMA scans to evaluate for treatment failure. And this wasn't always even in comparison to an initial PSMA PET scan. It became really concerning because we're using a much more sensitive tool to assess for new disease, perhaps even in comparison to older imaging anyway, but we have no definitions of what progression looks like on a PSMA PET scan. So that's where this genesis came from.

In talking to some of the community docs about these issues, they were like, "I'm glad you said this because I wasn't comfortable with this. My partner did it and I wasn't sure." So we thought we'd put this out there to at least raise the flag that I think we may be using a very important new technology that has a clearly defined role in newly diagnosed and even recurrent patients, but we're using it, perhaps, in ways in metastatic castration-resistant disease that are detrimental to patients.

The reason why I fear it's detrimental is really that when an imaging radiologist person is looking at the scans, it's very easy for them to define what they call progression, which sometimes it's new lesions on a PSMA PET, sometimes it's higher SUVs, which I think are completely unknown what that even means. But none of these parameters have been used to define treatment failures of any of the therapies we use in our clinics in prostate cancer. And I think that's the important distinction to make, is that progression, as defined by a radiologist, has value on a report, but what we need in clinic as medical oncologists is evidence of treatment failure. And evidence of treatment failure requires rigorous testing and prospective trials. And frankly, that's how our clinics have existed now.

Very much so, Prostate Cancer Working Group criteria with CAT scans and bone scans have given us the template that allowed for innumerable phase III trials over the last decade that have added a number of wonderful therapies for our patients. And we know what treatment failure looks like based on those prospective trials, but we have no idea what's going on with PET. And the concern is that we're using a more sensitive tool and we're stopping a therapy that's working based on established criteria prematurely, and in that way, maybe we're not really helping our patients out at all.

Oliver Sartor: Ravi, I'm really appreciating the fact that you're bringing this to life, so I think it's a very important question. Let me just editorialize for a brief moment that if you gather 15 or 20 experts in a room and ask, "What is disease progression as measured on a PSMA PET?" Well, you might have 15 or 20 different answers. There is no consensus. And even more importantly, lacking consensus is due to the fact that we lack data and we don't really have the proper way to assess.

Now, I give a little bit of a shoutout to the RECIP people, but nevertheless, never validated in a prospective manner, and particularly it could vary from treatment to treatment. So it's not just going to be a one-size-fits-all. A new lesion on a scan does not mean that the patient is suffering a detriment. And by the way, it may occur in the context of 15 lesions going away. So we have to take it in a totality of context, not just a single lesion in and of itself.

Let me ask you, I read your paper, of course, and there were a couple of comments that were a little bit interesting. One thing that you said was whether or not in micrometastatic disease that we might be detecting an indolent state that would be without progression for a period of time. Now, that's a very provocative comment, and I wonder if you could elaborate because I know you're doing trials and I know a little bit because you and I have talked. Tell me a little bit about this indolent metastatic disease state that you're referring to.

Ravi Madan: Yeah, so I think there's three levels to this. One, you refer to a prospective trial that we're conducting here in biochemically recurrent disease, and we're monitoring patients over, hopefully, years. And what we're seeing, that, despite rising PSAs, PSMA scans don't change that much. I think that these 3 or 4-millimeter lymph nodes that we're seeing in recurrent disease may not be clinically meaningful for years, and we're probably overtreating too many patients. But that's in the castration-sensitive setting.

I think in the castration-resistant setting, there's a couple of other points to make. One is that we conducted a study years ago with the old sodium fluoride PET, which was, in its day, going to change the world until it didn't. And then PSMA came along. But we conducted a study where we did serial sodium fluoride PET scans in patients with stable disease on enzalutamide. And it was remarkable the lesions that came and went in those patients. We saw lesions appear, the CAT scan and bone scan were stable so we left them on the therapy, and then the next scan they were gone. And so I think we need to realize that these are very sensitive imaging modalities. We might be catching things that are clinical snapshots of things that may not be pressing clinically over time.

There's an interesting citation, I think it was number 14 from a Hopkins group, but they actually did biopsies of PSMA-recurrent disease versus conventional imaging-recurrent disease, and the genomics underlying the PSMA-positive disease only was not as virulent or as aggressive, which, I think, is not surprising. So I think that, yes, we're going to see a lot more. I don't know if it's clinically relevant. We may not know which lesions are going to become clinically relevant. And I think the other thing that's tied to this is I think it's very unlikely that a patient with only a PSMA-positive finding, say in a bone with a negative technetium scan, is going to have symptoms from that lesion. That's a little more opinion probably on my part, but I think that, generally speaking, if you're going to have a destructive process that's going to be symptomatic, you'd think you'd probably pick it up on a technetium as well. I actually would like your thoughts on that since I have an expert.

Oliver Sartor: No, no, I agree completely, Ravi. This is going to be typically micrometastatic disease. It's going to be subclinical in almost all, and since I can't even think of a single case where we had a PSMA PET-positive conventional imaging-negative lesion cause symptoms. It's just not going to happen.

Ravi Madan: I agree with you on that.

Oliver Sartor: Yeah. I really like this distinction between treatment failure and PET progression. I'd like to hear just a little bit more depth about that. And then if you don't mind, I'm going to ask you a challenging question. How are we going to determine eventually how these PSMA PET lesions might contribute to treatment failure? So not only today's perspective, but take us to tomorrow and ask the question, how are we going to really show that this is relevant or not?

Ravi Madan: Right. Right now in our clinic when we use conventional imaging, old school CAT scan and bone scan, we know that when we meet the criteria that were established in prospective phase III trials, that if we stop at that point, that the patients still have an overall survival benefit. And we know that from the way the trials were done. And so that is at least a somewhat validated form of treatment failure, that the treatment has failed, we should move on, but we've maximized the benefit. And I think maximizing the benefit is important in a disease where we don't have infinite therapies. If we had 10 therapies, 20 therapies we could run through in serial, this wouldn't be an issue. But the reality is there's kind of this false sense of being proactive by catching something early on a PET scan and thinking that we're going to move on to the next therapy. But most times in prostate cancer, the next therapy is more toxic and doesn't work for as long. And in that way, again, we're short-changing patients.

So I think that treatment failure has been defined in trials that have demonstrated a survival advantage. And we don't know what that looks like on PSMA PET when we're talking about a new lymph node or a new bone lesion. The way to know that, which is your challenging question, is challenging for the field. It requires hard work. It requires what I call the Rosetta Stone, the way to basically interpret CAT scans and bone scans and understand what it means on a PSMA PET. And the only way, really, that I can think of doing that is in current prospective trials where you do paired imaging of all three together, and then you'll have the data and the data will tell us the story. I just think that I get labeled maybe as a Luddite for writing something like this with my colleagues, but the truth is we have the data for the old technology, we need it for the new, and unless we're doing all of this together in parallel, we're not going to understand what that Rosetta Stone looks like.

Oliver Sartor: No, I think that's very well taken, Ravi, and the distinction between the past trials where we've demonstrated prolongation and survival with a particular approach toward imaging and now all of a sudden things are changing, it really represents a challenge for our field because we're finding, in the community as well as in trials, that there may be new lesions on a PSMA PET, and then all of a sudden things change.

Let me give you an exact example of that in Australia. You may or may not know Louise Emmett. I think she's terrific by the way, and put together a trial called ENZA-p, recently published in Lancet Oncology. She had really good PSA data, and it turned out that her rPFS data was significantly compromised. And so I asked Louise, "Why, Louise, the compromise on the rPFS?" She said, "Well, people would see a new lesion on a PSMA PET scan and immediately assume that it's progression and then not even get conventional imaging so we had to move on." Now, eventually they'll have survival on that trial, but the question will remain, were patients terminated too early and we don't even know?

I'm going to do one more question before we wrap up, and this is going to be another provocative question. I'm really interested in your answer. We do, at times, have patients with what I'm going to term oligo progression. Sometimes that's conventional imaging, sometimes it's PSMA PET. Do you carve that out as a special category? And if so, do you react to it in your clinic? Do you react to it in a trial? How do you manage what we might term oligo progression as detected on both conventional imaging and PSMA PET scans?

Ravi Madan: We stumbled upon this in that PET study that I alluded to earlier where we had first-line mCRPC patients treated with enzalutamide in parallel with other people who've kind of stumbled upon this as well. What we found is that progression is probably occurring in a minority of the lesions. We worked with a company called AIQ, you probably know them up in Madison, Wisconsin, and they used AI to determine which lesions were progressing. We surreptitiously, in one patient, radiated a patient whose PSA had climbed from 3 to 35, and then his PSA dropped to like 3 again and he was stable for another two years for a total of five years on enzalutamide. We have a prospective trial that's in the works that we're trying to get off the ground where we're looking specifically at using AI to define oligo progression.

But I think you are asking about, can PSMA be detecting this oligo progression and should we react to that? And I think we need to ask that question prospectively in trials. I think the challenge is to understand what are the lesions that need to be treated and what don't? And that's, again, where we need that serial data, where we're not overreacting to PSMA PET changes while conventional imaging stays stable. And what actually happens to those lesions? Do we need AI? Does PET scan tell us that?

But I think what that's also telling us, and the very concept of all of this, is the underlying therapy is still having an effect sometimes on the majority of the disease. So it may be possible for us to take a patient on first-line, say, enzalutamide or some other androgen receptor pathway inhibitor that's doing great, isolate an area of oligo progression, and sustain them on that first-line therapy for even longer. So the goal wouldn't be necessarily to detect oligo progression, and that could be done with radiation, that could be done with other therapies, other platforms, but the goal would not be to detect oligo progression and rapidly switch a therapy, in my opinion. I think the ideal thing for the patient is to eliminate that site that's breaking through the dam, if you will, and allow the underlying therapy that's stable with the rest of the disease to be continued months, hopefully years, longer.

Oliver Sartor: Ravi, this is a very provocative conversation. By the way, I agree with you again. The interplay between systemic therapies and focal therapies, whether it be radiation or whatever, is really an interesting area right now. We're migrating to new horizons because of the sensitivity of our scans, but I think that your article sort of asks us to have wisdom and not simply be reactive when a new spot comes on without understanding the context or the interpretation of that. And your important mandate stated in the paper was, "Be careful about changing your systemic therapies because you might be losing a valuable asset and moving on to a more toxic therapy that's not even required."

Ravi Madan: Can I make one more comment?

Oliver Sartor: Of course. Absolutely.

Ravi Madan: This is implicit, but I think it should be said explicitly. We need to also get the conventional imaging at baseline. So in patients who are newly diagnosed and they have all these findings on PSMA PET and we put them on treatment for mCSPC, in order to understand what the benchmark or baseline is years from now, we don't just need to get restaging scans with conventional imaging, the conventional imaging needs to be done at baseline as well. So we're always going to have to have that benchmark-to-benchmark comparison until we get new data. And I'm sorry to interrupt you, but I thought that was probably something that was implicit in all our discussion, but warranted some explicit mention.

Oliver Sartor: No, I agree, Ravi, thank you for making the point. Thank you for being here today. Thank you for looking at this in a critical fashion. Thank you for asking the questions that are important for our field and all the best in your studies. So thank you again for being here today.

Ravi Madan: Thanks for chatting. It was great.

Oliver Sartor: Thanks, Ravi. Bye.