Read the Full Video Transcript

Robert Reiter: So there will be a little bit of overlap from the last session. First of all, I want to, of course, thank Tom, Jeremie, and Johannes for putting this together. We did something similar when MRI just began about 20 years ago, 15 years ago, and it was really, really helpful, I think, to kind of educate the broader community about MRI, and I think this type of thing will really help to improve the use and utility of PSMA. So anyway, I'm going to talk about the clinical impact that PSMA had prior to prostatectomy. I think we all know this. The goals of imaging for localized disease include staging. We want to detect disease, detect metastatic disease. We want to know if it's multifocal. And what I'm going to talk a little bit later in this talk about is really kind of is there any functional or biological information that PSMA imaging can tell us to actually help in terms of our management decisions as well?

So the outline, I'm going to try and go through these hopefully fairly quickly. The first thing I'm going to talk about is a little bit like what we've already heard. So I'll go through this quickly, which is the utility of PSMA as an adjunct to MRI for the detection of prostate cancer. Then I'm going to talk about some data from UCLA and elsewhere about PSMA PET/CT as a predictor of outcome, as a predictor of biochemical recurrence. Then I'm going to talk about some new work by one of my fellows, Adam Weiner, who's looking really kind of at the exploration of what does PSMA tell us about tumor biology. And finally, kind of in line with what Peter showed with fluorescent imaging, I'll show a single case using radio-guided surgery for patients who have had metastatic disease at diagnosis rather than in the salvage setting.

So you saw this from Tom, so I won't go through this more. Obviously, PSMA is far superior to conventional imaging for the detection of particularly pelvic nodal disease. In a cohort who underwent surgery, so this isn't the true sensitivity of PSMA, PSMA had a 40% sensitivity, but I think the sensitivity for metastatic nodes overall, not including just surgical patients, is probably higher. But it does illustrate that we still need to do better in terms of detection of nodal disease to be able to use this optimally intraoperatively either with fluorescence or radio-guided surgery.

So this is going back to the days of MRI just a few years ago. This is a paper that we published showing that MRI is not perfect. This raised a lot of eyebrows because there were those in the world who said, "MRI will cure everything that ails you." But basically, about 20 to 40% of grade two cancers are missed by MRI. We went on later on to actually show that the tumors that were missed by MRI, genomically appeared to be less aggressive than those that were detected, particularly if you compare PIRADS-5 to PIRADS-2 lesions, suggesting that perhaps we don't need to detect everything in the prostate. And that's important I think when we add PSMA PET to it.

But just exploring the synergy or the additive information of PSMA PET and MRI. This was actually a paper from Louise and John Yaxley and others in Australia demonstrating that the combined sensitivity of PSMA and MRI is a little bit better than PSMA or MRI alone. But as you can see, an 80% sensitivity for PSMA and MRI independently is quite good. So I don't think that we're going to be using PSMA PET to augment MRI in every single patient, but I think it will be select patients as was discussed in the last discussion session. You saw some data from here too. This is a paper, this is a UCLA paper looking at the comparison of PSMA PET and MRI on whole mounts.

Now, these were patients who were in our prospective clinical trial. So generally, they had either unfavorable intermediate-risk disease or high-risk disease. So this did not include really any grade two patients. And in this series, the sensitivity of both PSMA PET and MRI was very high, 90%, and there really wasn't a whole lot of additive value in this high-risk population. Kind of going back to what I just said, that it may not be necessary in all patients, but I think, for instance, patients who have PIRADS-3, they're equivocal or they have a negative MRI, but you're highly suspicious, I think clearly there's some additive value to PSMA PET.

Since that paper, Idasony, who I think is here, but she has a poster, which I'm going to call out with the next few slides, has really been doing a much more in-depth analysis, basically mapping PSMA PET and MRI in a large series over a thousand whole mounts that we have. And she has some of that early data back there. And I'm just going to show one slide from that, which basically is asking, okay, so what's the difference between MRI positive PET positive lesions versus those where there may be discordance? And there are two findings that came out of that. Number one, the larger the tumor, the more likely it will be visible both by MRI and PSMA PET. Where there's discordance, where PSMA is positive and MRI is negative or vice versa, generally those are smaller tumors. And also ISUP grade, so higher-grade tumors are more likely to be detected. You've heard that theme already a little bit this morning. And where there's discordance is often times in these grade two or lower cancers where some are PSMA PET positive, MRI negative, and vice versa.

So the second thing I want to talk about is a little bit about, okay, so PSMA is a great imaging agent. We can localize disease, we can detect almost all of it, but what does it tell us about outcome? So this excerpt is a slide that should have been in the last session. This is just a case series from Europe where they took 45 men who had a negative MRI or couldn't get an MRI for whatever reason, and they did PSMA PET. And basically, it just shows that in cases that are MRI negative, but you have a high degree of suspicion, many of those do have cancer that can be detected by PSMA PET.

So this is a paper from Australia as well, just looking at SUVmax as a, whether or not it's a prognostic indicator of biochemical recurrence and basically shows if you divide PSMA expression into quartiles, that the highest quartile expression has a higher risk of biochemical recurrence. This kind of does correlate with the notion that PSMA uptake is associated or correlated with tumor aggressiveness and it's borne out. So clearly, PSMA may be telling us more than simply there is disease or there is not disease. It's probably telling us overall something about grade. And it definitely, there's an association with biochemical recurrence. And this is a study that was published, Jeremy and others at UCLA and Wolfgang, basically everybody who's here today. And this is again from that prospective study that was run in patients who had PSMA PET prior to surgery or at diagnosis and then underwent surgery. And basically, the question was, could the information from PSMA PET, particularly extraprostatic disease, so lymph nodes or elsewhere, could that improve the prediction or prognosis of patients at the time of diagnosis compared to traditional clinical pathologic variables such as a CAPRA score, which Peter Carroll mentioned. And how did that compare to postoperative information in terms of prediction of biochemical recurrence?

And basically, what that study showed, it's so small, I can barely see it, it's on in the upper left panel that the higher the expression of PSMA, the higher the uptake as in the study from Australia, higher risk of biochemical recurrence. So just the uptake in the primary tumor did provide some correlation with prediction of biochemical recurrence in univariate analysis. In panel B, if there was extraprostatic extension, i.e., lymph node positive, obviously those patients did much worse. So PSMA PET really added to our traditional variables in terms of predicting biochemical recurrence. C is using CAPRA, so basically high versus low risk, scores of less than five or greater than five. And basically, you can see that it performs well, but not as well as in panel B where you're using the PSMA PET information as well. And then I think that's the most important stuff from that slide.

And then if you actually do various models, and what you do is you ask what's the impact of PSMA PET in terms of extraprostatic extension, when you add it to scores such as CAPRA and the high PSMA uptake in the prostate, you can see it really adds significant prognostic information. It kind of makes sense if there's extraprostatic disease, the outcomes will be worse. But clearly, this is information that we didn't have prior to the advent of PSMA PET.

So the next question is whether PSMA... So it clearly is telling us something about prognosis. What does it tell us about tumor biology? And this is a case kind of like what we saw before. This is a patient who had actually very high-grade disease, and you can see he's got this tumor on the left that actually is Gleason grade five, but has absolutely no PSMA uptake. And he had a positive lymph node as well. Also no PSMA uptake. So clearly, there are PSMA-negative tumors. And the question is, what is that? Is that telling us something about biology or about management?

So this is in a poster by Adam Weiner back there. I think it's poster number five. I'll talk more about it later. But basically, what he's done is he started to look at what pathways, what signaling pathways may be regulated or related to PSMA. He used the TCGA database. He looked at PSMA expression. He first was able to show that PSMA expression does correlate reasonably well with PSMA uptake on imaging. So it's a reasonable surrogate to look at. And basically, you can see that there are pathways associated with very high PSMA expression. It's regulated by androgen we know. So it's clearly associated with androgen response pathways, cell growth, metabolism, whereas PSMA low is actually associated with very adverse features such as inflammatory pathways, hypoxia, epithelial-mesenchymal transition, and stemness, which is associated with lineage plasticity. But this is looking just at primary tumors themselves.

Here he shows that in adenocarcinoma, which is generally AR regulated, there's very high expression of PSMA. Whereas in neuroendocrine prostate cancer, tumors that have undergone lineage plasticity, there's lower expression. Again, that correlates with this idea of AR-regulated versus more stem-like and de-differentiated. And then clinically, what does that mean? He looked at a number of prospective databases that were out there and he asked, okay, so if a tumor is low PSMA versus high PSMA, does that predict outcome related to the use of androgen deprivation therapy? And in this series, he could show that patients who had very low PSMA expression tended to do worse with androgen deprivation compared to those who had higher expression because those tumors may be more dependent on AR signaling.

This is just showing the association of PSMA with stemness. So low PSMA abundance is associated with features of stem cells or EMT in prostate cancer. And so what does that tell us? Perhaps it says something about the response to radiation. Tumors that are low in PSMA or more neuroendocrine often have p53 mutations, are RB loss, which are associated with lineage plasticity and associated with radiation resistance. Hypoxia is also associated with radiation resistance. And so this series too is a prospective trial. He looked at PSMA expression. And again, PSMA low predicted for a worse outcome with radiation that may be compensated by higher doses of radiation or ADT addition, but nevertheless, suggesting again that PSMA may be adding to our knowledge biologically of what's going on with the tumor.

And for the final minute or two, I'll talk about a single case. This is kind of as a follow-up to what Peter showed with fluorescent-guided imaging. Since I didn't have a fluorescent probe, we decided to do radio-guided surgery. So basically, we take a PSMA probe, label it with technetium. This is done by all my great colleagues at UCLA, Jeremy and Magnus, and many others. And then basically, we take the patient to surgery the next day and we use a gamma probe to detect that signal intraoperatively in patients who we suspect have nodal metastasis. And why do we need to do this? This shows us why, because as surgeons, even though we think we know what we're doing, if I see a left obturator node on a PSMA PET and I go in blindly and I take out all the left obturator nodes, I will miss disease half the time. So we're not very good at localizing disease, even with three-dimensional imaging; we really need something intraoperatively, whether it's fluorescence or radio guidance to do it.

And this is how it works. The day before surgery, the patient gets injected with Technetium 99 PSMA I&S, takes them to surgery. We've got a really good probe that we can use intraoperatively. It's a drop-in probe with a robot to see if we can detect and resect the disease. And I'm just going to show one case. So this is a patient who maybe historically I would not have operated on if I knew he had lymph node metastasis. But he's a super young guy. He had a PSA of 96. He had very high-grade disease. And on his PSMA PET, he actually had what looked like two nodes. He clearly had a right, we thought it was an obturator node, but actually, it's probably a very deep right internal iliac node. He also had a very small left external iliac node. He decided to be managed on a Proteus-like protocol where he got six months of neoadjuvant ADT and abiraterone, and then six months after and had surgery after six months.

He did very well, but we did image him prior to surgery. He still had uptake in a single lymph node on the right side, and we proceeded to take him to surgery knowing that he may well need radiation. So this is his PET showing this deep internal iliac node. After therapy for six months of ADT, you could see his prostate lit up because he had such high-risk disease. We then took him to the operating room, and let's see if this will play. So this is the gamma probe on the right that we've got. And basically, this is where we think the node is located. And on the bottom, you can see kind of the gamma signal. So now you can see kind of where we found it right there. It's actually really perivascular, basically right on the surface of the bladder. A place like Peter said, I would never go without knowing that there was actually something there for risk of injuring the bladder.

And this is basically the dissection. You can see how medial this is located. But the point is that we were able to go in, we were able to identify that lymph node, completely remove it, remove the rest of the lymph nodes on that side. Also did the left side. And basically, what we found is there was one lymph node positive on the right side. That was the node that we were after. And there was still a small node on the left side that the signal had gone away on PET but was persistent, and he's doing very well postoperatively. So I'll finish there.

PSMA PET can improve the detection of cancer when added to MRI and vice versa. It's a predictor of biochemical recurrence and may improve upon standard preoperative factors. It may truly provide important biological information to use in the future to assess treatment and make treatment decisions. And radio guidance, like radiofluorescent surgery, may improve the resection of extraprostatic disease. And as Peter showed, I've been interested in fluorescent imaging for many, many years. We've done a lot of preclinical work and I really think it is the future of surgery. Thank you very much.