Lecture Presentation:

Phillip Koo: Welcome to the Precision Medicine Forum here at UroToday, where we are focusing on PSMA. We're very fortunate to have with us today two international authorities on the topic, and as we sort of develop this program, we're going to start off by looking at the diagnostic applications of the various PSMA targeted radioisotopes that are currently in development, and then we'll move on to the PSMA targeted therapies. So to kick things off, we have Dr. Thomas Hope who's the Director of Molecular Therapy at the University of California, San Francisco. Thank you very much, Dr. Hope, for joining us today.

Thomas Hope: Great. So I will give a short background on the use of PSMA PET imaging in prostate cancer to sort of give us a foundation before we get into our discussion. So as a background, PSMA stands for prostate-specific membrane antigen, and it's a transmembrane protein that's overexpressed in prostate cancer. It's been used numerous times before in particular with [inaudible 00:01:20]. It was used for imaging and it was not fairly useful because [inaudible 00:01:24] bound to the intracellular component of the molecule. But the development of small molecules was based on a urea back on the bind to the extracellular component of the molecule has really brought this molecule back into the forefront. One of the odd things about it is that PSMA actually isn't specific to prostate cancer. It's also, overexpressed in a number of other tumors and other benign etiologies. We won't go into that too much today, but it's important to know when you're interpreting these imaging studies moving forward.

So I'll just start with just one molecule, gallium 68 PSMA-11. I think it's useful to think about how these molecules are designed so that you can think about the variety of different compounds that are out there. There's hundreds of these PSMA targeted molecules and so I want to put them in perspective. So PSMA-11 has two components. One component on this side of the molecule outlined in blue as a chelator. In this case, it's an HBED-CC chelator which binds to a radio metal. And you can use in this case, gallium 68 for imaging. On the other side of the molecule, is this targeting motif. It's a urea motif that binds to the PSMA molecule and gives the sensitivity and specificity of this molecule for PSMA.

Now we take gallium 68, put it into the chelator, and inject that into humans. And that gives us images like this. So this is a 69-year-old man whose status post radical prostatectomy. You can see his PSA is 0.7. And in those images, you can see both the right humeral metastasis, the right arm there, as well as the left internal iliac nodal metastasis. And I like this case because it highlights both how well PSMA PET can localize both soft tissue disease as well as osseous disease in the same imaging study. And our hope moving forward is that instead of getting different imaging studies for bones and for soft tissues and the prostate bed, we'll move towards having a single imaging study that answers all of these questions at the same time.

Now, as I said before, there's a number of molecules that target PSMA. And if you look at this, there's a number that are labeled with technetium that can be images of SPECT molecules that are labeled with radiometals, such as gallium, as well as molecules labeled with fluorine. But as you can see across the majority of these molecules, there's a commonality that they all have this urea motif, which gives that specificity to targeting to PSMA. So I think of these as a class of compounds. Now, if you look at how these images, whole-body images from some of those compounds look, they're very similar. You get salivary gland uptake, renal excretion, but there are some differences. So for example, on the left here, we see PSMA-11 and then PSMA-1007. PSMA-1007 is different and there's much more hepatobiliary clearance. So you don't see a lot of bladder activity, which can help you visualize local disease, which might be potentially obscured based on that bladder activity. DCFPyL is another compound that's finished enrollment in Phase III trials, which actually is very similar to PSMA-11 and bio-distribution is likely very similar in terms of ability to detect disease.

There's other compounds like CTT1057 which I put up because it comes from my institution, which have a slightly different bio-distribution. You can see a little bit more blood pool activity if the one hour time point compared to the urea-based molecules. Overall though, I consider PSMA targeting agents to be within a class and function very similarly, although the bio-distributions might vary slightly. And there's a large number of these compounds currently in clinical development, the one closest to approval is PSMA-11 and soon thereafter we anticipate the DCFPyL will be approved by the FDA as well, hopefully, middle to late next year.

Now, if we go back to what got us all interested in PSMA was these early articles comparing PSMA and fluorocholine. We don't think about it much in the United States, but outside of the United States fluorocholine was a standard imaging modality for patients with biochemical recurrent prostate cancer. And if you compare PSMA to fluorocholine, PSMA outperforms it. On the left side is the first article at the University of Heidelberg, which compared PSMA-11 to fluorocholine and showed us a detection rate. And this is at a patient level, how many patients did you actually localize disease in? But note that that study had an average PSA of 11. And the second study, which came out of St. Vincent's Hospital in Sydney, you can see they imaged one more patient, which I always think is humorous. I always promised to do a study with 39 patients in it.

But in that study, they had a lower PSA, a PSA of 1.7. And in these patients with lower PSA, you can start to see the detection sensitivity between PSMA and fluorocholine starting to separate. So it's really in patients with low PSAs where the detectability increases with PSMA compared to other radiotracers and this is in those patients where you potentially have the ability to cure patients because they might have oligometastatic disease at this earlier stage of cancer.

Now in the United States, the other FDA approved radiotracer for localizing prostate cancer is fluciclovine. And so here's a case comparing fluciclovine and PSMA PET in a patient who was imaged one week apart. And you can see that in both fluciclovine and PSMA, you can actually see this right pelvic sidewall metastasis, but the tumor to background ratio is much, much better with PSMA compared to fluciclovine. It's only about three with fluciclovine versus over 50 with PSMA. And it's that increase in targeted background ratio which really improves our ability to localize disease and feel more confident that we're actually detecting the tumor that's present. In a different patient, you can see in this patient who has right seminal vesicle recurrence, that the fluciclovine has very mild uptake in the right seminal vesicle, but it's much more apparent on the PSMA PET. And even though there is more bladder activity because of the renal excretion, I'm much more confident calling that right seminal vascular recurrence than I would be with fluciclovine PET.

Now, Johannes Czernin who's going to talk next was put together as trial Jeremie Calias at UCLA, which was a wonderful trial comparing PSMA PET to fluciclovine PET in patients after radical prostatectomy with a PSA less than two. So that patient population, which is very hard to detect recurrent disease and overall, and you can see the detection rate with PSMA was two-fold higher, greater than that with fluciclovine and I should add that all of these studies were read by three blinded readers and it was very robustly performed. I think one of the additions to that increase in detection rate of two-fold, if you look at the location of disease, it's very interesting. So for example, the nodal disease was 30% with PSMA. 30% of those 50 patients had nodal disease, but only 8% with FACBC or fluciclovine. Whereas it's actually the local tumor where nearly all patients that had disease localized with fluciclovine was only in the prostate itself. Fluciclovine in essence had no metastatic disease in these patients. So I think it's really important to realize that PSMA, particularly in patients with low PSA outperforms these other imaging modalities, such as fluorocholine and fluciclovine.

Now in collaboration with UCLA, with Dr. Czernin we ran a Phase III trial using PSMA-11 PET to localize disease and patients with biochemical recurrence. And this data serves as the background for the two NDAs that we collaboratively submitted to the FDA. And you can see, we used this composite endpoint to localize disease and determine whether or not they are true or false positives, which resulted in a positive predictive value of 0.92. In the chart on the right, you can see here, the different detection rates based on the PSMA level. So between a PSA of 1 to 2.0, we had a detection rate of over 80%, but when the PSA went below 0.5, the detection rate fell below 40%. And you got to remember that this study was done with blinded readers. So in order to detect disease, the blind readers had to agree with each other. In literature that uses a single reader that's not blind to the image study will always have a bias to having a higher detection rate than studies that are performed with blinded readers. So keep that in mind when you're comparing detection rates across studies.

I think the most important point of this study, this inter-reader variability. The inter-reader variability was very good in the PSMA-11 with a kappa ranging around 0.65 to 0.78, but an average around a 0.7, which is very good for an imaging study. If you look at a kappa for example in the study published by Calais with fluciclovine, the kappa is lower around a range of 0.2, and that's because it's much more difficult to interpret a fluciclovine PET compared to a PSMA PET.

Now, the really key here is trying to determine whether or not a patient has oligometastatic disease. Like you see in this patient whose PSA after radical prostatectomy went from 0.5 to 0.9. And if you see oligometastatic disease like this, the hypothesis that we can target it with external beam radiation therapy, and potentially cure the patient. Now we at UCSF went back and looked at a cohort of patients who had a PSA less than two after radical prostatectomy. And then we looked at consensus tumor volumes based on standard radiotherapy planning and noted that in the patients who had PSMA avid disease, 30% of patients out of the 125 would have had disease missed by standard consensus, tumor volume, radiation therapy. And that is really important because if you think about the percent of patients who fail salvage radiotherapy after radical prostatectomy, it's about 30%. And so it really suggests that the failure of salvage radiation therapy is not because of radiation therapy itself failing, but because of the fact that we cannot localize disease prior to treatment planning and therefore the disease is missed by the radiation therapy fields.

The other thing we've looked at is how PSMA PET impacts the management of patients. And not surprisingly, there's a large impact of PSMA PET overall. In nearly every study that's been published shows a greater than 50% major change in management and patients. And I like this chart because it really shows how the management is dependent on the results and the PSMA PET. So for example, if a patient has no evidence of disease on the PSMA PET, the majority of patients end up converting to getting active surveillance or surveillance. If patients have nodal disease, then patients tend to be getting radiation therapy to the prostate and/or the pelvic nodes. And in patients who have metastatic disease, obviously, there's a dramatic change towards patients receiving systemic therapies, such as abiraterone or other systemic drugs.

So overall there's three main points. So one, PSMA PET is superior to existing radiotracers for the detection of metastatic prostate cancer. I think it's really important to realize that the use of PSMA PET has a large impact on clinical care in particular with radiation therapy planning. And then lastly, gallium PSMA-11 NDA is currently under review by the FDA, and that was performed collaterally between UCLA and UCSF and we expect soon DCFPyL will be submitting an NDA, and we're looking forward to the approval of both those drugs. So thank you very much.

Faculty Discussion:

Phillip Koo: Great. Thank you, Dr. Hope, for that great summary of the clinical applications of PSMA diagnostic agents. So next we have Dr. Czernin, who's the head of the Ahmanson Translational Imaging and Therapy Center, UCLA, and also the Editor in Chief of the Journal of Nuclear Medicine who is going to speak about the use of PSMA as a therapeutic target in prostate cancers. Thank you, Dr. Czernin for joining us.

Johannes Czernin: Thank you very much for inviting me. I'm going to focus on the PSMA as a relevant therapeutic target. And I'm almost exclusively talking about prospective clinical trials because I think there are not sufficient data available to really come to more definitive conclusions about the efficacy and the side effect profile of these therapeutics.

So here are my disclosures. Tom showed already this kind of schematic in a slightly different way. I just want to make the point here that this [inaudible 00:13:54] of course can be labeled with gallium or F18, this would be PSMA-11, this would be the fluorinated compounds like PYL, but they also can be labeled to a treatment with beta emitters such as lutetium-117 or alpha emitters such as 225 actinium. Because there are no real data out on the effectiveness and efficacy of a actinium labeled PSMA, I will limit discussion pretty much on lutetium labeled compounds. Just to make the point very clear that alpha emitters have a theoretical advantage in terms of the energy transplant, the curative energy deposition. It deposits higher energy than lutetium and the radiation effect would be a likely more profound to some of the targets, especially if there may be smaller targets. But again, the actinium is not established yet then there are no prospective trials really investigating the performance of these compounds.

This was the first landmark study and it comes from Michael Hofman from Peter MacCallum in Melbourne. This is a study that really raised the attention very much to the potential of lutetium PSMA as an effective radionuclide treatment in patients with very advanced disease. What you see in these images are pre and post-therapy images where red denotes tumor involvement, and you always have one, a maximum intensity projection image that is done before the treatment and the one next to it after treatment. You see the dramatic differences between is pre and post-therapy scans in pretty much all of these patients here. They over reported a PSA decline by 50% in 57% of all patients. And this is a very attractive image. So it became the Society of Nuclear Medicine and Molecular Imaging image of the year in 2019.

And the values of PSA reduction are very dramatic here. These little numbers are PSA levels. Some PSA levels can go from close 1000 to seven, from 200 to 0.4 and from 148 to 1.6, dramatic reductions and they're very high response rates. And I will later on talk about the fact that the group in Australia has actually improved patient selection for including into lutetium PSMA trials in prostate cancer, which probably accounts for the higher response rates.

I also want to point to the side effect profile, and maybe just focusing these 30 patients that were published here on the grade 4 toxicity. You see that the grade 4 toxicity is very low with three patients with thrombocytopenia, but no other grade 4 side effects. This treatment is extremely well tolerated by almost all patients.

They recently published in a follow-up study with an expanded cohort of 50 patients you again see the PSA response rate. When you take 50% decrease in PSA as your response threshold, you see that 64% had such a decrease, but very interestingly, pretty much all patients had some PSA decrease in response to treatment, which again must be due to their patient selection. They also did an assessment of pain response. This is a pain severity and pain interference index shown here. And it all cycles up to two, three, four, and three months follow-up. For both of these parameters, there's a significant reduction in bone pain and improved pain interference. This was published recently in the Journal of Nuclear Medicine.

And then there's some survival data here. This is the survival for the entire cohort. The median survival was 13.3 months. This is the PSA progression-free survival, which was around six months, and here the survival stratified by PSA responders versus non-responders patient who responded with greater 50% reduction and then 18.4-month survival versus the non-responders at an 8.7. So a significant improvement in survival when patients have at least 50% decrease in PSA.

There's a second interesting component here and that is that 15 of their patients were retreated. So they had responses, but after a certain time, PSA started rising again, these patients underwent re-treatment, and again they achieved 73% of patients with a PSA decline of 50%. And here you show the patients are still alive. And so you get even with retreating these patients, some really durable responses so that the median overall survival from start of the initial treatment to the endpoint was 26.6 months overall.

We did a bicentric study. This was headed initially by [inaudible 00:19:26] in Finland and now by Jeremie Calais at UCLA and [inaudible 00:19:32] at an entity called Excel Diagnostics in Houston. There was a prospective open-label, single-arm Phase II clinical trial, and we enrolled together more than 60 patients in 2017. We started and ended the enrollment right about exactly two years ago. The reason for stopping the trial was that the Phase III VISION trial started and that, of course there was no great interest in competing enrollment between the two trials.

But we do have some interesting data that have not been published yet. It was actually a randomized trial. We randomized two administered activities, 6.0 versus 7.4 gigabecquerel. And these are the data. These are the numbers that we get. We got a 50% decrease in PSA after two cycles, which was the prospective endpoint in only 27% of the patients. The best PSA response in 38% of patients with 50% decline and any PSA decline in 61% of the patients. It is interesting that 10 of the 64 patients are still alive and under active surveillance some of them with PSA levels that are still undetectable. So some of these patients that are also referred to as super responders sometimes, and we don't know why they are super responders, can have exceptional longterm outcomes with this treatment.

This is now a study that was done by a coworker of us now, but it originated really in Munich, it's by Andrei Gafita. We looked into various registries and databases in the Technical University of Munich, Peter MacCallum, UCLA, University of Houston, Heidelberg and Excel Diagnostics. In the screen, patients who underwent lutetium treatment, screened out those patients who had super scan appearance and ended up with about 40 patients, that the motivation for this was that usually patients with very advanced bone disease or bone involvement are excluded from these prospective clinical trials. And Andrei really wanted to know is that justified and what kind of responses can you get in these patients was recently published in European Urology.

So the questions that I ask is was there high-risk for bone marrow toxicity? And the second one of course is efficacy. These are all the patients involved, this is think, a fairly nice panel, at the top you see the manipulated maximum intensity projection images. There's a scaling of course, different between patients, but you see that there is extensive bone involvement in all of these patients, the threshold inclusion was 50% of total bone involvement. And in red, you see the threshold that kind of carves out all the bone involvement in these patients. You see that there's very extensive bone disease. So these are patients that we consider PSMA super scans.

And these are the results, their responses and side effects, because side effects in these patients were of high interest to us, of course and to Andrei. Forty-three patients, they had a median PSA of 1000, much higher than in other studies that were published. They did on average four cycles, similar to Hofman and patients that had equal to more than four cycles, fairly similar number too. They had very nice PSA responses. Sixty-five percent of these patients had reductions of 50%. The overall survival was 11.6 months.

The other thing that was important is the side effect profile, which is not different from, for instance, the Hofman study that didn't include patient with super scan. I also want to point to the difference between prospective and retrospective studies. Look at the safety profile here. When you do retrospective studies, it's impossible to accurately reflect what is really happening in patients with regards to side effects. So I think the importance of prospective studies cannot be questioned.

This is now a study that was conducted by Michael Hoffman, who is the leader of the Australia New Zealand trial that I showed here and who has done a large number and is doing a large number of prospective trials that will answer many, many highly relevant questions. Here they compared in a randomized fashion, the activity and safety of lutetium PSMA versus cabazitaxel. The eligibility criteria shown here, they had docetaxel, they were suitable of course for cabazitaxel with progressive disease with rising PSA, they had an adequate renal function and an ECOG performance status of zero to two. And then from the PSMA PET point of view, the PSMA has to be max of 20 at any site measurable sites with SUV max 10, nor FDG positive, but PSMA negative sites in individual patients and everything was centrally revealed, so very nicely designed study. Then they were randomized to getting lutetium PSMA-617 up to six cycles whereas cabazitaxel using this protocol here. And then it did expect to see after 24 hours, if there was an exceptional response, they would recommend discontinuation of treatment until progression occurs.

And this is what I find kind of a key aspect of the Australian Hofman approach if we'll call it, and that is to do FDG imaging in every patient. So look at the patients, 28% were ineligible, these 80 patients. When [inaudible 00:25:46] is low PSMA activity in these lesions, it's not negative, but low and highly positive FDG findings. So the tumor does not express the target, therefore treating these patients doesn't make sense. Here we have mixed pattern with PSMA positive lesions, but some of the lesions are only FDG positive and not PSA positive such as for instance, in the liver. And again, that's probably not going to get you very good responses, so much of the target is not expressed in so many lesions. And here the typical patients that would be included either with concurrent FDG PSMA positive the target is expressed in FDG negative scans.

So here now are the results. The key result is shown here, the best PSA response on the right side. When you take the lutetium PSMA data, 66% responders, more than 50% decrease in PSA, and some with a smaller decreases in PSA versus the cabazitaxel response with PSA response rate, again, defined as [inaudible 00:27:00] 50% in only 37% of patients. And this is here the proportion for event-free survival. In this case, this is a PSA progression-free survival, which was significantly longer for the lutetium PSMA group than for the cabazitaxel group. I want to mention this study was presented by Michael at the 2020 ASCO Virtual Meeting and the data still preliminary. It's not the final results because the study is continuing, but I think the data is highly promising and very promising for further utilization of lutetium PSMA. This is the safety profile and again, just to make this long story short, fewer side effects with lutetium PSMA than with cabazitaxel.

Finally, we all wait to hear the results of the people that face VISION trial, that as you all know, is a multicenter randomized Phase III trial in metastatic castrate-resistant prostate cancer that involves more than 750 patients, it's a two to one randomization. And it compares standard of care with intervention plus standard of care. And we of course all hope that the data will come out soon because a positive result of these trials will give the whole field of diagnostics a dramatic boost.

In summary, we get a very high PSA response rate to lutetium PSMA, especially if patients are more carefully selected using FDG PET in combination with PSMA imaging. We still don't know whether and how deep the overall survival benefit is, but the VISION trial will provide the answer. I think that the bone scan data, bone involvement data that come from [inaudible 00:29:01] multicenter trial suggests that one should include super scan patients in further studies and the extensive Hofman data strongly suggests that PSMA FDG mismatches need to be excluded if FDG has a disproportionately higher shown uptake than PSMA. But I don't think that that's the stop or the end of the story. I think it'd be to understand resistance mechanisms much better. There are still 50% non-responders and we clearly need rational combination therapies and need to position this in close collaboration with urology, uro-oncology community, to place lutetium therapy into the appropriate clinical position or management position.

I also feel very strongly that re-treatment studies need to be done. There's more than sufficient evidence now from several sites, several places that re-treatment actually is effective and the latest data point came from Michael Hofman here. And as a limitation of this presentation, I didn't have time to talk about actinium PSMA but maybe the subsequent discussion will give us an opportunity to do this. With this, I thank you very much for your attention.

Phillip Koo: Thank you very much Dr. Czernin for that great presentation. So Neal, I'll turn to you first. We've heard a lot about PSMA diagnostics and therapies over the past several years, but clearly we see the reality of this being available in the US very close. We could feel it. So what resonates most for the urology community from your standpoint?

Neal Shore: Well, you know, first of all, what phenomenal presentations just now by doctors Hope and Czernin. These are going to be remarkable changes in our imaging. And I think that urologists will now be able to get access and up to speed with the most accurate PET imaging scans that we've not been able to benefit or enjoy where we see them in parts of Europe, parts of Latin America and Australia, parts of Asia. I have to say thanks to UCLA and UCSF and doctors Hope and Czernin leading this NDA.

One of my questions is this is just such a great moment, and I guess we're hearing that or at least I've heard, maybe Tom can comment on this. We can expect to see the NDA hopefully before the end of 2020. And I guess the question is, and in keeping with what you're asking, Phil, is I've heard this in other conversation. Will PSMA PET ultimately replace the need for technetium bone scan and CT scan and during our evaluation? So that would be one of my first questions. Another question I have is if we're going to see the PYL, the sodium fluoride and the PSMA-11 gallium based scans sort of come to the US market at around the same time, are there any aspects to that processing and adoption, that community, medical oncology and urology clinics, will there be a differentiation? You did a really nice job talking about the bio-distribution differentials, but I'm just thinking in terms of practicality of getting them up and going for use?

Thomas Hope: Johannes I'll take this one first and then you can correct me when I'm wrong. So I think there's sort of two or three questions embedded in there. The first one was what happens in conventional imaging in the era of PSMA PET? That's a very good question, which the issue to a certain extent is a lot of trial data that determines how patients are treated are based on imaging findings. In particular, if there are more than three or four lesions, docetaxel high-volume/low-volume cutoffs, are all based on bone scan. And it's unclear how to translate those findings to a PSMA PET. So some work has to be done in order to do that translation, but at the same time, PSMA PET is superior in nearly every way to a bone scan and a CT scan.

So I think that as time moves forward, my anticipation is that will slowly fade away. I know that for example, at our institution, we have medical oncologists still order a bone scan in patients with metastatic disease seen on PSMA PET to see whether or not they would have been bone scan positive or negative, which is the way it works. So that's sort of the first question.

The second question about how does PSMA-11 DCFPyL play out in the market, I think I'll rephrase it that way. So let's take PSMA-11 first. So PSMA-11 is going to be approved by UCSF and UCLA this fall if the FDA can ever figure out how to inspect us in the environment of COVID. So assuming that actually happens which it should, it will be available at two institutions, not every institution, right? It will take one to two more years for companies to get NDA's or make kits that contain the compound and distribute it or have radiopharmacies such as PETNET or SOFIE, these are the companies that can actually make it and then make it locally. So it takes a while for that network to become in essence nationalized. At the same time, DCFPyL is a different story, right? It's a fluorinated compound, it's owned by a company it's going to be distributed. So in essence, within a few months after approval of DCFPyL, DCFPyL will be more widely available. It's easier to make and distribute around the country. So my anticipation is that sometime year when DCFPyL would come out, it will be much more widely available and used.

Then there's a whole, much more complicated discussion of reimbursement, which is probably not worth getting into here that will ploy into which compounds will people use, because there's actually, these things will end up being expensive. So that will actually make it more complicated. Does that help answer that?

Phillip Koo: Yeah, that helps tremendously. And we're entering to the era now where the nuclear medicine department, the nuclear medicine radiologist, become an integral part of the multidisciplinary team for treating prostate cancer. Whether you're a urologist or medical oncologist, which I think is great.

Johannes Czernin: I just want to add one more thing to the bone scan issue, because we looked at a large number of PSMA studies, several hundred and stratified by the indication that there is initial diagnosis, biochemistry, recurrence, re-staging, whatever you name it, and looked at the detection rates of bone involvement. And they are much, much higher and it's surprisingly low PSA levels to detect already quite a bit of bone involvement. So I think eventually all the bone scan criteria or bone met criteria will have to be rewritten. All the guidelines will have to be changed because you see absolutely much, much, much more at any level of PSA, not only the ones where the PSA is not that informative because of systemic treatments and so on, but in early patients.

So I think that will really change the management dramatically. And the other thing that Neal, you mentioned that I think is becoming part of the management team. That we see already because despite the fact that the PSMA scan is still expensive because it's done on the cost, patients have to pay for the scans essentially, we do now six patients a day and there's no tumor board at UCLA that leaves out PSMA image as a critical component of the best care of the patient. So I completely agree with you. The other thing that I found always interesting before I start talking too much, is that we never needed marketing for this PSMA thing. The demand was there from the first scans that we did. I think Tom has the same experience we never had to ask, "Oh, we have this great new scan and so on." No, it started that day one with incredible demand from physicians and patients, and it never stopped.

Phillip Koo: I think that's an important point that there is such a demand for these technologies out there. So, Alicia, you were part of the VISION trial and you know a lot of the excitement about the future of PSMA targeted therapies being real in the US. What are your thoughts from a medical oncologist perspective about the future of this type of therapy?

Alicia Morgans: Great, thank you. And yes, we were so pleased and are still involved in the VISION trial as everyone knows, although it's completed enrollment, we are still in follow-up until we meet at events and then probably there'll be some extended followup. So we are pleased and so grateful to be able to participate in that study. Similar, as we heard regarding the scans, there was no marketing that we had this study, but the patients poured in really just hoping to have access to a therapy that seems to be quite tolerable and at least on early studies seems to be effective. One of the challenges actually on the VISION trial that I would love to hear Dr. Czernin's thoughts on were that patients who had a super scan were actually excluded from the trial. And you alluded to this, I think in your talk, if you didn't overtly mention it and it was very encouraging actually for me to see the data presented on those patients who had super scans, not actually having the catastrophic events that we anticipated may occur.

I'd love to hear you elaborate a little bit on that from a nuclear medicine perspective, because those are some of the patients who need the most help and it's hard to keep them out of these kinds of trials. It will be hard to not want to treat them should this drug be approved.

Johannes Czernin: Yeah. So I think the reason why there's much less toxicity than was expected, the bone-marrow toxicity because the way you could maybe envision it as liberating the bone marrow. And we really didn't see in this study, Andrei didn't see many terrible events. All the grade 4 events recovered so it was not irreversible bone marrow damage. There were no fatalities or anything like this, even not even close to that. So I think it's extremely well-tolerated. It also is quite effective, maybe much more effective in giving pain relief than 223-radium, which is in my view, not that effective in my experience. So I think it's going to be very useful. You saw the survival was 11 months, which is of course shorter than the survival in the PSA responders of other groups. But I think it's not a bad number and clearly there shouldn't be a reason to exclude them.

So that's one group of patients. I think that we need to focus on, that's the very sick ones but of course the other studies, and I'm sure they will be done in Australia very soon as moving it earlier and see whether you can actually do almost in high-risk patients post-surgical treatment, given that the side effects are so benign because we don't have that long followup. But I think one should probably mention, look at that too.

Alicia Morgans: I completely agree and I'm excited for that. There's actually an Alliance AFT trial here in the US that hopefully will be launching in the near future, looking at lutetium in the hormone-sensitive metastatic setting, which I think we're all also very, very excited about.

But as I think about these agents rolling out the other challenge with the VISION trial that I as a medical oncologist noticed and wanted to speak to Tom about was that in this trial, actually patients had to travel to a center where they could undergo the screening PSMA PET and with the PYL agent at least it seems like that's something that could be distributed and could be used. And the question I have is really kind of related to some of the things that you mentioned in your talk about reader reliability and it seemed like these PSMA agents, at least the one that you presented, which I think was the gallium agents the 11. So it seemed like those are better in terms of that inter-reader reliability and something that people can train relatively quickly to use and to use broadly.

And so as I just think about these agents going out into the world and the imaging also going out into the world, what are your thoughts on how well, this is something that could be implemented more broadly in these centers where they'd have to learn the technology that you've now had access to for a few years.

Thomas Hope: Yeah. So let me try to corral that question. The interpretation of it obviously depends on the setting of the disease. So in the setting of a VISION trial, the inter-reader reliability to determine whether or not there's high volume disease, that needs therapy would be like one. It's very easy to interpret those studies. As you get down to very low PSAs obviously the disagreement starts to increase. So obviously reader performance depends a lot on the setting that you're interpreting it in.

Now, that being said nuclear medicine decisions are fairly good at seeing things that are hot. So the difficulty, I think in interpretation isn't actually in terms of seeing things that are hot, it's learning about the false positives. And we oftentimes don't talk about false positives on PSMA PET but there's certain benign bone lesions like hemangiomas, celiac ganglia, dorsal root ganglia, things like that that are oftentimes PSMA positive. So there's a little bit of a learning curve in terms of interpretation and knowing what is a false positive. I wouldn't even turn them false positive because it's normal bio-distribution really but they could be misinterpreted. And the same thing goes with any PET imaging study that we interpret as well in other settings. But I think as an imaging study, it's much easier to interpret compared to something like fluciclovine and so I would expect it to function very well.

Neal Shore: So nuclear medicine radiologists are very good at spotting things that are hot. They're also very good at putting the most technically complicated visually appealing slides together. And no doubt, the two of you have done that today. So I'm going to also challenge you with complicated questions because you're good at complicated things. So getting back to the work you presented, Johannes, by Hofman regarding mostly the TheraP trial. Now this whole notion around FDG and PSMA concordance or discordance. Now we didn't do that in the VISION trial. And we also participated in that we did not get FDG PET. There was a feeling that a CT scan cross-sectional imaging was adequate. So I want you to please address that as that wasn't part of the VISION study.

My second complicating question, because I know you guys are good at complicated questions and complicated slides is, would there ostensibly ever be a role for theranostic use in biochemical relapse patients who had avidity and do you see that as for my urology colleagues, a way to delay or potentially obviate the need for testosterone suppression?

Johannes Czernin: So maybe I start with the first one and then ask Tom to answer to the second one or we can both. So in most of the retrospective trials that were published before the first Hofman trial and also including our prospective Phase II trial, there was neither FDG included nor the CT mismatch. So you have a mass that doesn't have PSMA expression. So ours was probably kind of the minimal project for the time. At that time the common notion was 98% of patients or lesions are PSMA positive anyway, and there was not much sophistication in the thought process. One would have to ask Michael Hofman how often a CT scan alone would have ... That would be actually an interesting study because he has the data to look at it from the CT scan alone would have provided already the correct answer without the FDG information.

I think that in advanced disease FDG will be used at times anyway to monitor treatment because it's not ideal to monitor PSMA targeted treatment with PSMA targeted imaging. So a metabolic tracer might be a bit better intermediate endpoint biomarker to do this. So FDG may be coming anywhere on the CT component is there so one can look at it. My biggest surprise was in a way from the data that Michael showed where they excluded 80 patients at 27%, I think it was of the patient for the randomized treatment trial. My biggest surprise was that the number is so high because, in our population, I don't see it that often, but of course we try to look for it now, actually, it's pretty rare, depending on how you set your PSMA threshold, of course. But Tom, what's your experience with that?

Thomas Hope: So with the role of FDG, we don't obviously, I think first off it's a unique circumstance at the Peter MacCallum Institute where they have in essence access to do FDG PET whenever they want, they don't have to worry about insurance coverage or anything like that. So in the US we don't have a ton of experience with combining FDG with PSMA. we do have a trial where we're enrolling patients to try to learn some data about that. My general belief is that conventional imaging will catch them. So if you took Mike's data, the 28% that didn't qualify because they had the two imaging studies, 10% were because they had low PSMA uptake. So you can just take that out because that's obviously not going to be an issue and-

Johannes Czernin: But is that the number that you also see, the 10% with low PSMA? It's higher than what we see, I think.

Thomas Hope: We don't know because we don't image that many patients with it. I think, as you pointed out in your talk, Johannes, the rate of response in sites that aren't doing what Michael Hofman does is lower, right? The percent of patients who have PSA 50 response is lower. I think we all am assuming you as well are assuming that's because they're taking out these patients who are less likely to respond. That's our assumption. There might be other reasons for their higher response rate than that. So there's probably something to what they're doing. It's just unclear. It would be nice to have included them and then retrospectively been able to stratify based on those imaging studies to actually tell because the way those studies were designed, they weren't treated. And so you don't know if those were predictive in terms of response. I think that's just an assumption we're making.

At the same time, I think there's other ways to get at some of that data. For example, you can consider whole-body diffusion as a marker of cellularity and you wouldn't need to get an FDG. I think in the United States, we're going to struggle to get an FDG and a PSMA PET reimbursed within a two-week timeframe. I don't think that's ever happening. So we're going to be stuck just because of our insurance architecture at either getting conventional imaging, CT with contrast, or an MR in addition to PSMA PET. And I think that'll capture the majority of those patients or at least the ones that he's provided images. You'll see those little metastases. It'll be easy to see them on CT scan. So overall, I think that's where we're going to end up.

Johannes Czernin: And I think the second question that Neal asked us about essentially early treatment, I guess, that I'm of course, highly interested in. I mean, it would be theoretically, nice to convert prostate cancer to thyroid cancer and do the proposed surgical ablation therapy. I have no idea whether that's biologically ... We have anecdotal evidence and in some of our patients, we have fairly low disease burden. The responses are at times spectacular. And I wonder sometimes whether you have too much disease and no treatment can be lasting is successful. So this definitely needs to be done. Maybe Tom, we should start thinking about doing a study. It has to be done because we will never know before the prospective trial is done. Probably [inaudible 00:51:22] is doing it already I will bet.

Thomas Hope: I have some hesitation to a certain extent in that particularly with the lutetium based radioligand therapy. When you have a beta particle travel distance of like over a millimeter to treat lesions that are less than a millimeter in diameter, micrometastases, lutetium is never going to be effective in that setting. And so when you're talking about biochemical recurrence with a low PSA the radiation therapy, external beam is nice because you're hitting a field. So you're going to presumably deposit a therapeutic dose of radiation across anything that's within that field. And so really what the PSMA PET is telling you is saying, "Hey, there's localized disease in this area, aim at it." And I'm hesitant to start using lutetium in that setting. It's possible actinium could play a role there, but that's further out.

Johannes Czernin: But there would be, of course, those patients who as you showed from your data, they have more disease and not limited to the pelvis at the time of biochemically recurrent, which is often surprising that would not be candidates for actual external beam radiation, maybe potentially. All that needs to be done prospectively. Actually Michael Hofman wrote a very nice editorial in the Journal of Nuclear Medicine to kind of highlight the need for well-designed prospective studies in nuclear medicine, because for decades, I think we were suffering from underpowered kind of semi-prospective and pointless and therefore meaningless studies. And I think it's very important that we all kind of do something with Australia does so well at this, put together teams of people to create these prospective trials.

Phillip Koo: So I'm going to have to bring this to a close. Clearly, there's lots of excitement here. One thing that I am personally excited about, it's great to see nuclear medicine physicians talk about these trials and talk about being active members and not just talk, actually being an active part of the trial design, trial implementation, which is a huge step forward for our specialty and something that I think we need even more of. And clearly we could sit here and probably talk for two hours and maybe even longer if we had drinks, but thank you all for your time and your expertise. It's great and I thank you on behalf of all prostate cancer patients, because the two of you have been pivotal members in actually making this a reality in the United States. So thank you and we hope to speak with you again soon.

Johannes Czernin: Thank you very much.

Thomas Hope: Yeah. Thank you all.

Alicia Morgans: Thank you.