Read the Full Video Transcript

Neal Shore: Well, hi everybody and welcome to UroToday and our online medical education program entitled Beyond Androgen Blockade, New Pathways and Novel Treatments in MHSPC and MCRPC. I'm Neal Shore and I have the great privilege and honor pleasure today to moderate our discussion with my dear friend and colleague Oliver Sartor, who's going to present for you today his topic, PSMA PET as a biomarker for advanced prostate cancer. Thanks very much, Oliver.

Oliver Sartor: Thank you very much Neal, and really a pleasure to be here today. I have the opportunity to talk about PSMA PET as a biomarker in advanced prostate cancer. I think everybody knows a lot about PSMA today because of the PSMA PET scans that are changing our practice. But just a reminder, PSMA is a transmembrane protein. It's relatively selectively expressed in the prostate, not exclusively. But it is expressed on the prostate to a great degree, and that allows us to both image and treat.

Now, I think everybody, as stated, understands the phenomenal role that PSMA PET has been playing in the staging of patients with unfavorable intermediate risk, high risk, and I call it a localized with quotation prostate cancer. Because it used to be localized when we would use conventional imaging and DREs and MRIs to be able to stage those patients. But now of course we're finding there's a lot more to the story for a number of these folks.

That's not only true for that high risk localized, but also for those with a PSMA recurrence after definitive therapies such as radical prostatectomy and radiation. That's not what I'm going to talk about today however. I'm going to talk about PSMA as a biomarker in advanced disease. It turns out that we've examined this biomarker in a context of a phase three trial called the VISION trial. Just briefly, I'm going to run over VISION and then get down to the heart of the talk.

VISION is this very important trial looking at the Lutetium-177-PSMA-617. I'm just going to call that Lutetium from now on just to keep it short. This was randomized trial looking at standard of care plus or minus Lutetium with two to one randomization and looking at rPFS and OS. Importantly, these patients had requirements not only for their pre-treatments, they had to have an inter receptor pathway inhibitor and progress.

They had to have a taxane, at least one had progressed, but they were selected on the basis of their PSMA PET positivity. Here's how we did it. We looked at PSMA PET positivity as being an uptake greater than the liver, and then we excluded patients who would have PSMA PET negative metastatic lesions. Here's the actual breakdown. People may not have seen this before. But when we did the scanning on the patient, by the way, 87% of the total patients actually enrolled after the scan.
But 4.9% of the patients that were screened with a PSMA PET actually did not have a positive lesion at all, so they were excluded. Then there were patients who had a positive lesion, but the individuals were excluded because they ended up having a lymph node of greater than 2.5 centimeters so it was PSMA PET negative. That meant uptake less than liver. Or a solid organ metastatic lesion and one centimeter or greater that was PSMA PET negative.

Or the soft tissue component of a bone metastatic lesion that was more than one centimeter and PSMA PET negative. This led to an exclusion of 8.7%. A total of 8.7 were excluded because they had PSMA PET negative metastasis. 4.9 were excluded because they had no evidence of PET positive disease. That's the way the selection took place. The rPFS published in New England Journal positive. We've all seen this. The OS positive, we've seen this.

Clearly we can say that using these criteria that there was a positive outcome for the treatment with Lutetium. Now let's dig into this a little bit more. There was an FDA required exploratory post-hoc analysis of the patient using quantitative PSMA PET. What they wanted to do is to look at quantitative parameters from the baseline PET and how that related to outcome. Now, Phil Koo presented this at the EANM, European Association Nuclear Medicine meeting this last fall.

This is a very recent, and many people may not have seen that. There are a bunch of authors on the manuscript, same authors that were on the VISION manuscript. When they did a whole variety of analyses, they came up with the tumor SUV mean, the whole body SUV mean, being the most important parameter from the imaging. Now, there's a little bit of unsatisfaction with this. Let me tell you what it is. This is a calculated value and not typically available.

We're working now on some nomograms that are going to be a little bit easier, and I'll introduce those over the latter part of the talk. But it turned out that the whole body tumor SUV mean was the one that was most associated with either rPFS or death. Now, here's the actual data and we've divided the whole body SUV mean on the baseline PET into quartiles. You can see on the left-hand side you have the Lutetium arm. You can look at the medians 5.8, 7.8, 9.8, 14.1.

When you have more PSMA PET uptake, you do better. That's because you have more of the target and it means you can bind the Lutetium more effectively to the tumor. That's the way I interpret it. Now for the standard of care arm in the control, you could see that there really was no real relationship between the quartile of the PSMA PET uptake and the outcome by RRP FS 3.9, 2.5, 4.3, 3.9. It turns out that when you look at the curves, you can see more effect and more differences between the arms when you have the higher SUV.

I don't think it's a surprise, but it's nice to see it done quantitatively. Now, here is the data that is for overall survival, this hidden rPFS, this overall survival. It's interesting. When we looked at it in quartiles, the Lutetium arm, it wasn't quite as dramatic until you got that high SUV mean, and that was up about 9.9 on the SUV mean, and there was a distinct difference. You look at the quartiles, it's 14.5, 12.6, 14.6, and then 21.4.

Again in the control arm really not a lot of differences. It made a difference for the Lutetium treated patients against you're targeting the PSMA on the cell surface, but not so much for the control arm. This is a little bit complex, but I'm going to have you focus on the two left panels. In the upper part, you've actually got an rPFS, in the lower part you've got an OS. Now what you see is an interesting phenomenon. There's no real cut point.

It's a continuous relationship. Here is in this lower left, hopefully you can see that pointer, where you're looking at the SUV mean cutoff and then the hazard ratios and lower is better. You can see it's a continuous curve from starting with an SUV mean, a cut point of less or greater than seven. Then you come all the way through. It's like the higher you go the better it gets. But you can't say, oh, don't treat because it's below nine or don't treat because it's below eight because you're still having a positive effect.

Now this second little bracket here is patients without event. I think the easiest way to look at it, just pick out one of these, and I think this is the red one right here. These are patients who have an event when they're greater than the cut point. You can see at the lowest cut points, they're the most events. But the higher you go on the SUV mean cut point, you diminish the rate of events. That is because these patients are not having events.

If you have a very high SUV mean, it means that you don't really relapse very much and you don't die very much. This just shows the relationship between the baseline PET and the actual outcomes for rPFS and OS. Over here, you begin to look at the median survival by the hazard ratios. Again, without going into too much detail, I'll just simply say that if you've got a little higher cut point, you do a little better, but you can't draw a threshold on.

Now, there's one more piece of data, and this was presented by Ken Herrmann, also part of that VISION team at ASCO 2023. Because looking at PET is only one parameter among many that might be important. Now, there are three different columns here. There's overall survival, rPFS and PSA. At the top, probably not shown very well in this slide, but here you have that whole body SUV mean, and the longer the line on a nomogram, the more important it is.

You can see that whether it's OS, rPFS or PSA 50, that each of these parameters are predicted by the baseline PSMA PET. But when it comes to survival, there are a lot of other factors as well. Are you using opioids for pain? Is your ALT elevated? What's your hemoglobin? What is your absolute lymphocyte count? Et cetera, et cetera. These nomograms are now under preparation for a manuscript. We'll get a good look at it later.

In summary, I'm going to say the PSMA PET is a useful biomarker when choosing patients for PSMA-617 Lutetium therapy. In general, if you have more uptake, you're going to get a better response on the SUV mean. There's no clear cut point. You can't just say, well above five, it's good, below five, it's bad. You simply are going to have a greater probability for those with the more uptake, but there's no clear cut point. Then there are other variables that are important, especially for overall survival of the vision setting. By the way, this may or may not apply, but we move earlier in settings like PSMAfore. Neal, I hope this has been a little bit educational for you and happy to have a little discussion.

Neal Shore: Well, thanks so much, Oliver. First, just let me say that your pioneering work in radiopharmaceuticals, starting back with radium and now with Lutetium and many other radiopharmaceuticals. In addition, the work you're doing in improving our understanding of imaging, not only in conjunction with our radiation oncology colleagues, but also our nuclear medicine radiology colleagues globally, is an incredible boon to optimizing patient care.

Congratulations on your recent award you received at Astro. Very well deserved. Let me ask you, Oliver, one of the things I hear from many of my colleagues in the community is somewhat of the challenges that they're undergoing in calculation of SUV amongst nuclear medicine community radiologists. What are we doing to better organize this, educate upon this? I know you're doing a lot, but the data you just presented today has arguably tremendous predictive and even potentially prognostic value, which will help us with our therapeutic selections with radiopharmaceuticals, specifically an RLT.

Oliver Sartor: Yeah, great question, Neal. By the way, I think the frustration that you made note of is true within some of the academic communities, not just the community setting. I'm going to say that this is being addressed in several different ways, and I will say that the nuclear medicine societies like SNMMI and EANM are taking a lead. It probably is going to require some type of artificial intelligence and software. There is a company, I'll call them out by name, Exini, that I've done a little work with.

They're attempting to look at this in a quantitative artificial intelligence enhanced manner using software that could be implemented in a cloud-based way available to pretty much anybody anywhere. It's making progress, but there are others working on it too. I happen to know that group, and they've built out some pretty damn nice software. I'll simply say that we're going to have to keep our hats on a little bit longer while waiting for this progress. But it's going to come. It's going to come. It's just not quite here right now.

Neal Shore: Yeah, I appreciate that. I think that's going to give us additional accuracy and understanding. As you've heralded clearly, newer imaging technologies, PSMA PET revolutionizing what we're doing. PSMA radioligand therapy revolutionizing how we're taking care of these patients. By the way, congratulations on your presidential address at ESMO this year on the PSMAfore, looking at Lutetium specifically in the pre chemotherapy mCRPC arena. One final question, Oliver, some of our colleagues in different parts of the world have a greater access and maybe there's some cost incentive or de-incentivizations as we have in the US regarding the use of FDG PET. What are your thoughts? I know you've given a lot of effort and analysis regarding other cross-sectional imaging. How do you recommend now regarding FDG PET in the setting of PSMA PET?

Oliver Sartor: Yeah, great question, Neal. Just for a tiny bit of background. The Australians have clearly shown the utility of what I'll call the double PET approach using both the FDG and the PSMA. But we were designing the VISION trial. It was acknowledged by all that if we put two different PET scans as an inclusion criteria and that one of those PET scans would probably not be reimbursed, that we would've created a huge hurdle for the adoption of this therapy into the community.

That's why we use the combination of cross-sectional imaging in one PET, the PSMA PET as opposed to two. Now, having said from the practical perspective of trying to get two PETs approved and paid for, it's just hard to do. But there is utility, in my opinion, in the FDG PET, and in some ways it's a little bit easier than using the cross-sectional imaging because you see the spots in a positive way rather than defining the PSMA PET negative on cross-sectional.

I will say that both will work. What we did in VISION will work, it was a nice study by UCLA where they treated pretty much everybody, and then they applied retrospectively the VISION criteria, and guess what? The VISION criteria patients did better than those that did not meet the VISION criteria. Nevertheless, FTG PET, I think is an important potential tool, but it's difficult to access in the United States because of the reimbursement issues regarding getting two PETs instead of one. If I had to choose one, I'd use a PSMA PET, not the FTG.

Neal Shore: Well, thank you so much. You bring up so many other interesting aspects to this. With any new novel imaging technology or new therapeutics, there's costs associated. We're always trying to create greater access throughout our countries that where we practice and of course globally. That's an important consideration. But you can't get there unless you have wonderful clinical outcome data that you've been leading the charge on, both from an imaging standpoint and a therapeutic standpoint. Professor Oliver Sartor, thank you so much for being part of the masterclass and really appreciate the work you've done and the work you're currently doing. Thanks very much.

Oliver Sartor: Thank you for having me, Neal. Appreciate it.