PSMA Imaging and Theranostics: Future Impact to Patient Management– Robert Flavell

August 15, 2018

(Length of presentation: 30 min)

In an interview with Phil Koo, Robert Flavell describes diagnostic and therapeutic applications of new agents in the management of patients with prostate cancer with initial high-risk diagnosis and biochemical recurrence. 


Robert R. Flavell, MD, Ph.D., Assistant Professor in Residence Department of Radiology, Section of Nuclear Medicine, University of California, San Francisco

Phillip J. Koo, MD, FACS

Further Related Content


Read the full video transcript

Dr. Phillip Koo: Hi, this is Phillip Koo from Banner MD Anderson Cancer Center in Phoenix, Arizona. Today, we're honored to have with us Dr. Robert Flavell, who is Assistant Professor in Residence in the Department of Radiology at UCSF in San Francisco. Dr. Flavell is the recipient of multiple junior investigative grants for multiple organizations such as the RSNA and the Prostate Cancer Research Foundation regarding prostate cancer imaging and theranostics and is really a burgeoning star within the field of nuclear medicine and radiology. We're very fortunate to have him with us today to talk about PSMA theranostics and how they might change practice in the future. 

Dr. Robert Flavell: Thanks very much. I'm really excited to share some of the recent innovations in this field on PSMA imaging and theranostics. Before I get started, I'd like to point out that this website also has a really nice and somewhat longer review on the historical development and state of the art on PSMA theranostics by Dr. Steve Cho. If you're interested in more information and especially historical perspective, ProstaScint, antibodies, things like that, I just probably recommend watching that video because it's really great. 

This lecture today is going to be focused on more recent developments and perhaps on how I and other people in the field anticipate that PSMA theranostics might actually be used in a clinical practice. Of course, none of this is really set in stone yet because this technology is pretty new and none of it is yet FDA approved. But, I think we all agree that this is a very exciting time for our field in nuclear medicine and molecular imaging and I think, there's a lot to offer PSMA based theranostics.

I have a few key points that I'd like to boil down to my main message. The first is that PSMA based PET imaging agents are changing the management of patients with prostate cancer and especially in the setting of patients with initial high risk diagnosis of prostate cancer and also in patients with biochemical recurrence after definitive therapy. Secondly, I would argue that PSMA agents are probably the most PET agents under development. The statement will be somewhat controversial but I'll provide some evidence to back up this claim. Many questions remain but a major role in the future will probably be guidance of metastasis directed therapy in patients with biochemical recurrence. Another important area will probably be PSMA based therapeutics, particularly in patients with castrate resistant prostate cancer.

If you look at kind of the historical development of PET imaging and prostate cancer, there's really been a tremendous amount of work done over the last couple of decades and if you look at this review for example, you're going to find a dozen different tracers out there. It's really pretty complicated and it's not entirely clear which one of these offers what. There's been a lot of comparisons but today, we're going to really focus on the PSMA based agents which is over here on the left side, here schematized as an antibody but we got more recent agents. 

You may ask yourself looking at this slide, "Well, you know, which one of these scans would I choose?" Well, it's really not yet entirely clear based on the literature so this is not yet a totally solved question. I would say that all of these PET based agents including choline, fluciclovine, also called Axumin and PSMA are superior to the prior standard of care, which would be CT scan and bone scan. There are small case series suggesting that fluciclovine and PSMA are superior to choline and there's a very small recent case here suggesting PSMA is superior to fluciclovine.

These are small cases but I think that they're based on my own personal experience and looking at these scans, they kind of make sense, so I'll say that this is my opinion but in my opinion for detecting a prostate cancer, CT scan or bone scan is worse than choline which in turn is worse than fluciclovine,  which is worse than PSMA but PSMA is not yet routinely available. There's a lot of other agents out there also that are at an earlier stage of development which may be complementing it as well. But for now, in my opinion, PSMA based agents are the most sensitive in specific agents for detecting prostate cancer although they are not yet FDA approved. 

Let's talk a little bit about what PSMA is. It's a transmembrane protein, it's highly expressed on prostate cancer cells at all stages and importantly, it's extra highly expressed in high-grade tumors, tumors that are androgen independent and metastatic tumors. In other words, your really bad actors, the ones you're the most worried about. It is also expressed on non-prostatic solid tumor neovascular and for this reason, a lot of people are now investigating PSMA in other tumors as well. There are lower level expression of this protein in healthy prostate, kidney, liver, small intestine and brain. 

This slide just kind of reviews the progress over the last decade in terms of small molecule development, a development of small molecule binders of PSMA. The work before that was previously done with the antibodies such as ProstaScint and J591 and if you're interested in that development, I strongly recommend checking out Steve Cho's lecture because I saw he really did a nice job reviewing that. 

More recently, there's been focus on these small molecular peptidomimetic base methods. These are mostly built around a couple different cores. The most commonly used is this urea based core here shown in blue. More recently, this phosphoramidite core has also shown some promise that included in some work from some of my colleagues here at UCSF. 

I'll just point out that most of these are designed around the same urea core and there, include three parts. There's the part that binds the PSMA, there's some kind of a linking moiety and then there's something out here that binds the radio label. If you look at these different agents that have been developed, they include a wide variety of different radio labels. It's really outside the scope of this talk to talk about the minute differences between these different tracers. They all have slightly different binding affinities, they have different radio isotopes but I'll just highlight some of the key differences that you can use in considering these different PSMA tracers. 

The first and most important is diagnostic versus therapeutic. The diagnostic isotopes of course, being use for diagnostic imaging. The therapeutic ones can label, essentially the same molecules as the diagnostic one, but now, you're just putting a therapeutic isotope, which will typically be the beta emitter or an alpha emitter. 

In terms of the diagnostic imaging options, the two main choices are PET versus SPECT, with the majority of the work being done in PET based on the improved resolution of that technique and also the fact that it's increasingly widely adopted in routine oncologic imaging. Among these tracers, you have various isotopes of choice and the two clear leaders in this area for the diagnostic side of things are F-18, which is really widely used in our country because FDG and really worldwide for that matter, so the F-18 widely available. Gallium-68 has, over the recent years, become more popular primarily because of the early work in neuroendocrine tumors and for that reason, a lot of the PSMA work has also been done using Gallium-68 base methods. 

Then, the other distinction is the core binding moiety, the part that binds the PSMA. As I said already, the most commonly used of this is the urea, although the phosphoramidite scaffold also shows some significant promise. 

Let's talk about what these scans look like. This is, here, one of the Gallium-68 PSMA-11 scans. This is probably the most widely used of the PSMA based diagnostic tracers. We've scanned hundreds of patients here at UCSF and before that, there was already hundreds of patients scanned in Europe and in Australia. When I'm teaching my residents about this scan, which now we're reading on a daily basis in the reading room, I tell them that they can usually recognize it by this so called scary clown appearance. I don't know if you see it. Not everyone buys it but it's an easy way to remember it because what you see is this bright uptake in these lacrimal glands and salivary glands. Do you see intense uptakes in spleen, kidneys and bladder and then lower level clearance by liver and the small bowel.

Let's just take a couple look at some scans that are abnormal. Here's a patient, again, similar distribution but now we see these other bright spots over here in the region of the right pelvis. These are metastases to the right ilium. Take a look at another one. This is a patient, again, has that distribution looks pretty similar other than now, we're seeing multiple spots in the retroperitoneum. I'll just point out to you that this patient's prostate bed is clear and the lower pelvis is also clear. This is a type of distribution we will commonly see in patients who have previously undergone salvage radiation. 

In other words, burned out most of the tumor down in the pelvis but maybe there was some above that radiation field and now, that's what we're seeing on the scan. Here's an example of an unfortunate patient who is diffused on metastasis. This was actually a patient with a new diagnosis of prostate cancer. You can see just a few uptake throughout the axial and appendicular skeleton here. 

Just to show you some example of what is the increased sensitivity of this study. Here's an example of a bone scan. This is a patient with already known metastatic cancer. This bone scan is not entirely normal. You can see, for example, there's uptake here in the left ilium and some sites in the spine. You would certainly not interpret this bone scan as normal although it is not horribly abnormal. We can have the CT scan through the lumbar spine. Not really a whole lot to see. Typical look for your prostate cancer with lots of degenerative changes. However, if you look at this patient's PSMA scan, you can see really, just the skeleton is just essentially riddled with metastatic disease. Almost all of which is occult on CT scan and the bone scan. 

This is what we see on a daily basis in the reading room is that we're just picking up a tremendous number of sites of disease that we're not picking up with their modalities. I'll tell you that our radiologist here, at UCSF, are really surprised when they see this. The common thing we actually get is what have we been missing all these years. 

What the question is, so knowing that we have this increased sensitivity scan, what do we actually do with this information? One thing is that you can use it for sensitive detection of metastatic disease in initial high risk staging. You can use it to detect sites of disease in biochemical recurrence, so patients who've have had rising PSA after having definitive therapy and really, the true clinical benefit of using this for guiding your patient’s management is not yet clear because it's so new. But there is really preliminary small study out there showing that it may cause delay and need for androgen deprivation therapy in biochemical recurrence population. 

The reason for this is that you can use the information from these scans to do so called metastasis directed therapy. In other words, these small sites of disease that are being picked up, you can hit them with targeted radiation or remove them surgically. 

Let's talk a little bit about the evidence supporting the use in new diagnosis. Here's an example of a study of 130 patients with newly diagnosed intermediate to high risk prostate cancer. This was a retrospective study and the imaging features were correlated with pathology and of these patients, 41 out of 130 patients had lymph node metastases, so about 32% and the sensitivity, specificity and accuracy of the gallium-PSMA was 66%, 99% and 89%.

You may say to yourself, "Whoa, 66%. That's really not that impressive." I think that's an error critique. It isn't more sensitive than the other techniques out there and what are they missing here, it's really tiny lymph nodes. These are one or two millimeter lymph nodes with small deposits of cancer in them. Know that   you are picking up most of the sites of disease. There are very small lymph node metastases and there's range of 1 or 2 millimeters in size that you will still miss with this technique. Although, it is clearly still superior to the other imaging techniques out there. 

It's probably the most sensitive imaging method for detecting metastatic disease in newly diagnosed or intermediate or high risk prostate cancer. The sensitivity is still moderate and there are several studies out there but they seem to be about 65% when correlated with surgical pathology. The role in changing management in the study and benefit remains unclear, although probably improved roles here would include guiding radiation planning and identifying distant metastatic disease. One trend I've noticed recently is that, for example, our radiation oncologists are interested in us in annotating all these lymph nodes so they can help guide focused radiation to several small lymph nodes. 

Let's give an example of how a scan like this could change a patient's management in the biochemical recurrence setting. This is a patient who underwent a prostatectomy in 2001 and salvage radiation in 2002 for biochemical recurrence. In 2001 had SBRT to an acetabular lesion and again in 2013 SBRT to an L2 lesion. He subsequently had another rise in PSA of 0.31 from undetectable in 2014 and this is what this patient's scan showed. There were two sites of metastatic disease. The top one here, you can see manifests as a small sclerotic lesion in the vertebral body, has increased uptake, the PSMA based agent. I think you could argue you might pick this up on a CT scan. This other site, there's no way you would ever would've picked it up on a CT scan. There's a tiny little lymph node here. Maybe a millimeter too in short axis dimension, increased uptake. In this case, the PSMA scan caused change in management, so this patient got radiation to these lesions. 

I'll just summarize a little bit of the data and how it changes management. This was a study performed here at UCSF by one of my colleagues, Tom Hope. There's 150 patients with biochemical recurrence of prostate cancer who underwent a PSMA PET. The referring physicians answered a survey regarding what would be the plan management without a PSMA study. Of these patients, 103 out of 150 had disease detected on the PSMA PET, so significant proportion and of these patients, 53% had a major change in management based on PSMA study. Recently, there were similar results published in a larger Australian study which I cited on the bottom there. 

Here's another example, 117 patients with hormone naïve biochemical recurrence and the detection rate by PSA was 0.02 to 0.5, 65%. Even at these very low PSA values, the detection rate is still quite high. Once you get above 0.5, you can be pretty confident, you're going to find something on the scan. 86% and it just gets better if you get higher. 86% from 1 to 2 and for greater 2 PSA, sensitivity was 100%. This is very sensitive for detecting metastatic disease in the biochemical recurrence setting. Really, what we've seen empirically is once you get above 0.5, you have a pretty good chance of picking something up.

Where are these sites of disease being found? Well, some of them are very much expected. For example, pelvic, retroperitoneal lymph nodes but sometimes you find surprises like mediastinal lymph  only or lungs only. These are ones that would've easily been missed using conventional imaging because they would've been so unexpected. 

Then the next question is, so yes, we can image these lesions, yes, we can treat them but does it matter? The role of this metastasis directed therapy in this study on the increased sensitivity of this newer scan is not yet totally clear and this is a matter of ongoing study and there are ongoing clinical trials for this purpose. There are early indicators that these may matter and delay a time for need to androgen deprivation therapy and in particular, I'd like to draw attention to this STOMP trial, which indicates the feasibility of metastasis directed therapy. 

Here's the data from this trial and this was a small trial where patients were randomized, surveillance versus metastatic dedicated therapy and in this case, they used choline PET. As I discussed earlier, choline PET is less sensitive and specific than PSMA. I think, probably, in future studies, with PSMA based agents, the results will be superior and then patients who had sites of disease identified on the choline scan had therapy with surgery or SBRT.

I'll just point out that the time to androgen deprivation initiation was longer in the patients who had metastases directed therapy. The P-value of this study did not achieve a statistical significance threshold but I think it shows a non-statistically significant trend and certainly forms the basis for future larger studies. 

The reason why these studies are going to take years to come out is because prostate cancer is a slower disease of course. If you look at the X-axis over here, these are several years in order to accumulate this data. This question is not going to be answered overnight. I am very optimistic that there will be a clinical benefit in this setting although it has yet to be proven. 

There are a few important details that I'd like to mention if you're ordering PSMA scans and if you're on the receiving end of these reports, which is that not everything that is hot is prostate cancer and we do see uptake in other benign and malignant processes and normal structures and the most notable of these ganglia and sympathetic ganglia and the celiac  access based on the cervical spine and the pelvis. The sensitivity is higher in high Gleason score and in rapid PSA doubling time and androgen deprivation therapy increases sensitivity acutely but decreases sensitivity in long-term use. 

I'll say that the patients where we see the most success is patients with rapid PSA doubling time, aggressive initial pathology and in those cases, we're picking up in the biochemical recurrence setting, patients with metastasis like with a PSA of 0.1 and things like that. Conversely, patients with very slow doubling times, patients that are on long-term androgen deprivation therapy, we will sometimes not find anything even at higher PSA values. 

What about PSMA based therapies? Many patients have been treated primarily with Lutetium-177 based methods in Australia and Europe. The initial data looked really good but there has been no phase three trials yet published although they are now open. Here's Lutetium and PSMA-617, which is the one with the most experience. This is a beta particle emitter, similar to the really classic radio iodine therapies and also similar to the recently FDA approved Lutetium-177 DOTATATE or Lutathera, which has been used for neuroendocrine tumors. 

These have been tested primarily in Europe and Australia, although they are now also being tested in this country and we're going to be opening here at UCSF hopefully in the next couple of months. 

What kind of data is available on these patients? As I said, there's not phase three data but what we have seen is that a significant percent of patients do have a response in the form of a PSA change and on imaging. 

This is a waterfall plot of a number of patients who received Lutetium-177 therapy and as you can see, a significant percentage of patients had a marked response in their PSA, although, clearly it's not everyone and there's significant population to patients that are also resistant to this therapy. 

I think another area of excitement is the use of other therapeutic isotopes and one that has attracted some attention is the Actinium-225, which is an alpha-emitter. Here's an example of a patient with spectacular response to the alpha emitter. The alpha emitters are perhaps superior because of the increased killing power of the alpha particle versus the beta particle. This is an example of the patient who had PSA of nearly 3,000, underwent 3 cycles of Actinium-PSMA and PSA dropped to nearly undetectable, one more cycle and it did drop to undetectable. Clearly, the spectacular response in this patient's case. 

Also, another interesting one. This is a patient who underwent cycle of Lutetium-PSMA, then had disease progression, so you can see that there's increase in extent of disease here, then underwent multiple cycles of Actinium-225 PSMA and had a complete response. 

There's a lot of excitement in this field about these therapies in general. We're looking forward to seeing then results of the phase three trials and accruing patients for that and the further back, but perhaps even more exciting in the future is the use of alpha emitters. 

I'm just going to conclude by summarizing the overall message, which is that PSMA based PET imaging agents are changing the management of patients with prostate cancer with initial high-risk diagnosis and biochemical recurrence, the PSMA agents are probably the most promising currently under development for PET imaging of prostate cancer although I'll give the caveat of that statement. It's controversial and people will disagree with it. There are many questions that remain but probably a major role will be the management of oligometastatic disease with metastasis directed therapy and the use of PSMA based therapeutics in castrate resistant prostate cancer.

I'd like to thank my colleagues Henry VanBrocklin and Tom Hope at UCSF for some of these slides and also I would like to thank you very much for your attention. 

Dr. Phillip Koo: Thank you, Robert, for that great presentation on PSMA. The diagnostic and therapeutic applications of these new agents. I wanted to ask a few questions as follow-up. What are your thoughts on these patients who don't express PSMA? There's literature out there that talks about 10 to 20% of patients actually being non-PSMA expressers. How do we handle those patients when it comes to the diagnostic piece and theranostic piece?

Dr. Robert Flavell: I think that's a really great question and one that hasn't been fully addressed by the studies yet out there. I will tell you just empirically, when I've seen these scans in real time, occasionally, you will see patients who come in with a new diagnosis of prostate cancer, have a decent sized tumor based on the biology or the known pathology and maybe a prostate MRI and we see no uptake anywhere on the scan.

How do we handle this? I think as a community, this is something that merits further study looking at these PSMA non-expressers and how can we best address them and what type of scan is best for this population but I think those of us who are looking at and interpreting these scans just recognize that there are some false negatives and we are reading these scans and we see patients that don't have PSMA expressing disease, it doesn't necessarily 100% mean that there's no tumor there and you have to pay attention of the other imaging and put it into the patient's whole context. 

Dr. Phillip Koo: Great. Thank you. All right, so another question. When we talk about these nuclear medicine therapies, I think there's a lot of baggage with nuclear medicine therapies with regards to burning out the bone marrow and side effects. When it comes to Lutetium based therapies, Lutetium-177, PSMA and Actinium, can you talk a little about the side effect profiles and what we can do minimize those side effects? 

Dr. Robert Flavell: I think that's a really great point. From what the early evidence on the PSMA based therapies has identified, two major sites of toxicity and the first is in the salivary glands, meaning they cause xerostomia. There's a sign that's a dry mouth and is something that is not unique to this type of therapy. In fact, there's a very well-known problem with radio iodine therapies as well. Maybe that this is something that we can help mitigate through some other pharmacologic agents or something like that. I haven't yet seen anything out there that's really addressed this definitively but I think it's a great area for future research. 

The second one and perhaps the more dangerous one is hitting the bone marrow and that's something we do see. Many of these patients have a heavy metastatic burden in the bone marrow and in many cases, will be pre-treated with bone marrow toxic chemotherapies, so we'll have limited bone marrow reserve to begin with. 

As to what we can do to mitigate that effect, I'm not really sure but it's something that we're going to have to pay close attention to and it's going to have to be considered in context in comparing with other therapies that also have bone marrow toxicity. For example, chemotherapy, radiation or radium-223 Xofigo. It's a real concern. It's something we're going to have to pay attention to and compare against the other therapies that are out there. 

Dr. Phillip Koo: Great. Last question, we read so much about PSMA. It's available widely in Europe, South America, Australia. What can providers in the US do to get access for their patients in the US?

Dr. Robert Flavell: That's a great question and I'll just repeat it real quick, which is that, basically, how do you get your patient in for a PSMA scan if they're interested in doing it? These scans are being provided primarily under two different mechanisms. The first is using a so-called cost recovery IND mechanism and that's how we're doing gallium-PSMA-11 scans here at UCSF. In that case, these are done using "open source technology" and they're being done simply to recoup the cost of the scan. These are offered at a relatively discounted price, although in most cases, these will not be covered by insurance, so the patient will be covering out of pocket. 

That is one option and the second main option is industry-sponsored clinical trials. Many of these agents are under development by companies and they are running various trials to investigate the sensitivity and specificity or utility of their agent. You can also try to get your patient on one of those trials. 

In terms of what site is offering it, we are offering the scans here at UCSF and if you're interested in getting your patient in for a scan here, then we can certainly arrange it. If you're looking at getting it done closer to home, you can just send me an email and I'd be happy to put you in touch with the closest PSMA center near to you. Don't hesitate to contact me via email. It's with any questions. 

Dr. Phillip Koo: Great. Thank you very much, Robert, for you time and this great presentation this morning.
email news signup