PSMA-targeted PET Imaging of Prostate Cancer - A Primer for Urologists - Michael Gorin

November 8, 2021

In this Society of Nuclear Medicine and Molecular Imaging (SNMMI), Satellite Symposium entitled PSMA-Targeted PET Imaging and Interpretation: What Urologists Need to Know Michael Gorin, MD presents on PSMA-targeted PET Imaging of Prostate Cancer - A Primer for Urologists. Dr. Gorin starts with reviewing conventional imaging modalities for prostate cancer which have proved insufficient for detecting prostate cancer, particularly in patients with low PSA values, and those patients with biochemically recurrent disease. PSMA-targeted PET has improved our ability to image prostate cancer, both in terms of sensitivity and specificity, across a range of disease states.


Michael Gorin, MD, Clinical Assistant Professor, Department of Urology, University of Pittsburgh School of Medicine, Pittsburgh, PA, Staff Urologist, UPMC Western Maryland, Cumberland, MD

Read the Full Video Transcript

Michael Gorin: PSMA-targeted PET Imaging of Prostate Cancer, a Primer for Urologists.

So first let's start with reviewing conventional imaging modalities for prostate cancer. Well, the audience should already be very familiar with imaging options for primary prostate cancer, and this includes a multi-parametric MRI of the prostate. This allows us to image the prostate gland itself and determine whether or not there are foci within it concerning the presence of prostate cancer. We also oftentimes use this imaging modality for imaging the pelvic lymph node basins to determine whether or not a patient has N1 disease. There are however significant limitations of this imaging modality, and that is that there is low specificity, relatively high expense, and significant issues with reproducibility, in particular when imaging the localized prostate itself. When imaging patients who have concern for metastatic prostate cancer, we by and large have relied for the last more than a decade on contrast-enhanced CT, as well as a bone scan. These imaging modalities also suffer from low sensitivity and with respect to a bone scan, also have issues with specificity in that oftentimes we see falsely positive findings.

So because of this, there has been a significant need in the field of urology and urologic oncology for imaging modalities that allow us to overcome the limitations of these conventional imaging techniques. And. this has been mostly solved through the development of molecular imaging techniques, most notably, which we are going to speak about today, PSMA-targeted PET imaging.

So what is molecular imaging? So I'll read you a description that has been published in the Journal of Radiology and put forth by the SNMMI for describing what actually molecular imaging is. So the term molecular imaging can be broadly defined as the in vivo characterization and measurement of biological processes at the cellular and molecular level. In contradistinction to "classical", what we would call anatomical, diagnostic imaging tests, it sets forth to probe the molecular abnormalities that are the basis of disease rather than to image the end effects of those molecular alterations. This by and large is typically performed using nuclear imaging techniques where a ligand is injected into a patient.  This ligand is radio-labeled, so it gives off some form of radiation, which can then be captured by either a PET or a SPECT scanner and allow the clinician to see the sites of disease.

So there is the potential for limitless sensitivity when one performs testing with molecular imaging. This is a schematic diagram to represent what I mean by that. Well, if we have a ligand with high affinity and it's been designed against a high-density target on cancer cells, we can, in theory, get enough radioactivity to even, just as small as one cell, to be able to see this using a molecular imaging test. In contrast, if we use a ligand that has a lower affinity or binds to a lower density target, it would take many, many more cancer cells, meaning at a larger volume, to be able to detect something with similar sensitivity.

So here, in this example, we have a tumor cell that has six radio-labeled ligands bound to it, which should, in theory, give the same amount of signal as this much, much larger tumor here that also is bound with six lower affinity or lower binding density radioligands. So there really is the limitless potential for sensitivity when using a molecular imaging test. In contrast, if one were to use conventional anatomic imaging, we would really require that tumor to grow to be large enough in size where it could be detected on the basis of saying something like a CT or an MRI.

In addition to that, there is the possibility for exceptional specificity with molecular imaging. If we design a radioligand that has a high target specificity, then we could really pick out tumors apart from other cells within the body that don't express the receptor for that ligand. So in this schematic example here, we could see a much higher degree of specificity with one of these ultra-specific ligands than if we had a ligand that binds not only to tumor cells but also to normal cells, as in the schematic here.

So the sort of mainstay of molecular imaging is performed with what is known as Positron Emission Tomography. So, urologists, I think, mostly think of a PET scan as sort of a black box and I hope to explain through this diagram here what is actually happening when we use a PET radiotracer. So PET radiotracers are labeled with positron emitters. Some examples of positron emitters are listed on the bottom of the slide here and these include things like F-18, gallium-68, C-11, copper-64, and I-124. So these radionuclides are bound to the ligand, whether that ligand is a small molecule or something like an antibody.  It then circulates around the body where it then binds to the receptor of interest. The positron emitter will emit a positron during a decay event. In this decay event, that positron will collide with a nearby electron. This nearby electron will then release two anti-parallel 511 keV photons. These anti-parallel photons, which are released, are then detected by the PET scanner and upon coincident detection of these photons, a signal will then be registered to allow for detection of the original binding event.

So what this allows for is a very, very specific detection and very, very clean images to be produced when imaging a patient with PET. This all stems from the fact that the anti-parallel released 511 keV photons must be detected by the PET scanner in order for a signal to be registered as real. So here, we have a tumor present in the stick figure patient, an annihilation event occurs where these two anti-parallel 511 keV photons have been detected by the rings of the PET scanner and this gets registered as a true detection event and is included in the final image.

There are a number of other naturally emitted 511 keV photons, what we call environmental photons, which are released by just our body, all different sorts of things in the environment. These are not produced as a result of an annihilation event, and so the rings of the detector will only see one of these photons rather than the anti-parallel event also be detected 180 degrees apart from it. And the PET scanner knows that this is an environmental photon and it actually subtracts this from the final image. So this allows us to have very, very clean images in the case of a PET scan, something that is not achieved with other molecular imaging modalities.

The molecular imaging modality that we oftentimes compare and contradistinction to PET scan is what is known as SPECT scan. So in SPECT, also known as single-photon emission tomography, we use photon emitters like technetium-99 or indium-111, and with these radiotracers, only a single photon is emitted and so the scanner picks up this photon, but it also has the potential to pick up similarly peaked wavelength photons, again, which are environmentally produced. So these images tend to have a bit more noise to them than what's possible with a PET scan.

So here, again, is a schematic representation of this and here we can see the true signal from the radiotracers included in the final image, but oftentimes also the lower level produced, but still present, environmental photons are also picked up by the SPECT scanner and become included in the final image. So because of that, these images tend to be, again, a bit noisier than PET.

So there are some other distinctions between PET and SPECT that I think are important for urologists to understand. So again, when we look at SPECT, we are using photon emitters, things like technetium or indium-111, whereas with PET we are using positron emitters. Again, the positron is released, causes an annihilation event, and what's actually detected is 511 keV photons by the PET scanner, but at the start of this process, we actually have a photon being emitted. Because of this, there's significantly much better spatial resolution inherent to SPECT. We could image things as small as 0.5 cm, and in truth, we could probably image things that are even smaller than this if we had radioligands that are able to deliver enough radionuclide density to something as small as one cell. In contrast, SPECT typically has a spatial resolution of only 1-2 cm. PET, for reasons which I won't go into in this talk, is intrinsically quantifiable, where SPECT is not. SPECT, however, offers the advantage that it is very low cost, and these scanners seem to be rather ubiquitous in hospitals throughout the world, whereas PET is a significantly higher cost technology and may not be available worldwide.

Because PET scan relies on the detection of only 511 keV photons, it is not dual tracer capable. No matter the radionuclide used with PET, you are still looking to detect that 511 keV photon, whereas with SPECT imaging, it's dual tracer compatible because we could detect a range of photon peaks and so you are able to actually administer something like technetium in combination with indium-111 and be able to detect a multiplex readout. So this is a potentially unique advantage of SPECT, not something that we have an application for yet in urology, but it is worth noting as we think about comparing the two technologies.

So there are a number of different molecular imaging options for prostate cancer. These are mostly PET radiotracers, but there is one notable SPECT radiotracer that I will cover. So the first radiotracer of interest is FDG PET. F-18-labeled FDG is the most commonly used radiotracer in the field of oncology imaging and the way that this radiotracer works is it is taken up by metabolically active cells, cells that have undergone aerobic glycolysis, the so-called Warburg effect. These rapidly dividing cells take up the FDG where it is then phosphorylated by hexokinase and gets trapped inside of the rapidly dividing cells. This is, again, used ubiquitously throughout the field of oncology, however, it doesn't really have much of a role in prostate cancer imaging and that is because prostate cancer cells by and large do not undergo the Warburg effect and so do not take up much in the way of FDG. And so because of that, we, as a field, have had to turn to other radiotracers.

So the next class of radiotracers I'd like to talk about, one being a SPECT radiotracer, and that being technetium-labeled methylene diphosphonate and the other being a PET radiotracer, which is sodium fluoride. These radiotracers work by binding to areas within the bone where it's turning over. So you're not actually imaging the cancer cells themselves, but what you are imaging is the end product of bone remodeling. This has served us well in the field of urologic oncology for imaging prostate cancer but does have the limitation of only being able to image bone and not soft tissue sites of disease.

So looking to overcome the issue with the fact that FDG PET relies on the Warburg effect, which doesn't typically exist in prostate cancer cells, and the shortcoming of methylene diphosphonate and sodium fluoride PET, that it only binds to bony sites of disease, there have been attempts to develop other radiotracers for prostate cancer imaging. Two which have been used for probably about the last decade now include C11-choline and F-18-labeled FACBC, which is actually known by its trade name now as Axumin.

These radiotracers have very similar mechanisms of action. They are taken up by rapidly dividing cells. In this case, it does not matter whether or not that is a prostate cancer cell or say an inflammatory cell that is undergoing division. C-11-choline is taken up because it's incorporated into the cell membranes of those rapidly dividing cells, whereas FACBC, again, also known as Axumin, is taken up by cells as an amino acid precursor, and allows for incorporation to newly made proteins. While these radiotracers do allow for the fairly sensitive detection of prostate cancer, they do suffer from the issue of specificity because their mechanism of uptake is not specific to prostate cancer and just simply that of a rapidly dividing cell.

So this then brings us to the PSMA-targeted radiotracers, which are really the focus of the rest of today's talk. The first radiotracer is known as gallium PSMA-11, or what simply has just been called in the literature as just gallium-PSMA, as well as DCFPyL, which is known by its trade name now as PYLARIFY® which is manufactured by Lantheus Pharmaceuticals. Both of these agents bind to PSMA. PSMA is a molecule on the cell surface of prostate cancer cells, and I will get into that more in a moment. It is then internalized into those cells and the PET radio signal is able to be detected. So this imaging agent is not only sensitive for prostate cancer, but it's also quite specific because it relies on the upregulation of the PSMA molecule, which is really confined to just prostate cancer, other malignancies yes, but in a given patient, will not be taken up by inflammatory cells and other rapidly dividing cells unless they express the PSMA molecule.

So over the last decade, there has been an absolute explosive growth in PSMA-targeted imaging. This is a search that I did of the PubMed database from 2010 to 2020 and you can really see the growth has almost been logarithmic in the number of publications, which is focused on PSMA imaging. I would say up until probably around 2019, most of this work was really limited to areas of the world where PSMA imaging was readily available, Australia and Germany. But now, as of late 2020, we now have access to two of these radiotracers, the PSMA-11, and the DCFPyL radiotracers, which we can now use clinically in our practice.

So what is PSMA? Well, it's a dimerized type II transmembrane glycoprotein. Under normal physiological conditions, it's responsible for the hydrolysis of N-acetylaspartylglutamate to glutamate, which is a neurotransmitter. So it is actually expressed normally by neurons, and that is how it was originally discovered in the neurological system, but what we then came to learn as a field is that it's actually, to a much higher extent, upregulated by prostate cancer cells. It is found in the neovasculature of a number of other solid malignancies, and [Dr. Rao 00:15:38] and myself have pushed forward with using this radiotracer for imaging renal cell carcinoma, which expresses PSMA in its neovasculature. And the truth is, we do not have a very good idea why it's expressed in the neovasculature of these cells, but suffice it to say, it is, making the name prostate-specific membrane antigen, a bit of a misnomer. But when speaking about prostate cancer imaging specifically, it is far more specific than those other mechanisms of binding that I had talked about earlier.

So again, PSMA is highly expressed by prostate cancer. Larger and more aggressive tumors tend to upregulate the highest levels of PSMA. Here, we can see some spots from a tissue microarray, where we can see Gleason 3+3=6 prostate cancer has very little PSMA expression, whereas, by the time we get to high-grade prostate cancers, the cancer cells are absolutely jam-packed with this molecule.

It is worth noting though, that PSMA expression can be lost in advanced stages of the disease. As cells undergo differentiation to a neuroendocrine-like phenotype, they tend to lose canonical androgen receptor signaling and with that, PSMA expression can be lost. But this is really only in cases of patients who see the selective pressure of agents which target androgen receptor signaling, so when we talk about patients who have never received androgen deprivation therapy, really, PSMA expression is highly, highly conserved and at very, very high levels. It's really only in the most advanced stages of the disease where we have to worry about this phenomenon and then the most advanced stage of disease, that's not really where we are using PSMA PET imaging clinically, as I'll go to in a moment. So while this issue is real and it is worth noting, I do not think it has particular clinical relevance, especially to urologists and really it only becomes an issue for medical oncologists.

So how can one target PSMA for both imaging and therapeutic purposes? Well, there are a number of ways one could do this. The most common way, at least in the modern era, is to design small molecule inhibitors which bind to the active site of the PSMA molecule. There are, however, antibodies that bind to the PSMA molecule, and these include the J591 antibody, as well as the 7E11 antibody. The 7E11 antibody is one that was incorporated previously into the ProstaScint radiotracer, which was a SPECT imaging agent, and I will get to that in just a few slides why that is no longer a terribly clinically relevant tool and some issues that have existed with it. Really, what we are talking about today, are these small molecule inhibitors that bind to the active site of the PSMA molecule.

So the most common way of binding to this active site is through urea-based small molecules. It turns out that urea binds to the active site of the PSMA molecule, and once you add to urea a linker and then a radionuclide, something like gallium or F-18, you could then have a PET radiotracer that binds to this molecule with very good affinity.

So many of the radiotracers that are available are built on this urea-based backbone. These are two radiotracers that were developed by mine and Steve's mentor, Dr. Martin Pomper at Johns Hopkins, and that includes DCFBC, as well as DCFPyL, DCFPyL being PYLARIFY, and the radiotracer that is now clinically available. We also have a handful of other radiotracers that suit the same purpose and are all built around the same chemistry, that chemistry, again, being a urea-based small molecule. Here is the PSMA-1007 molecule, and here is the rhPSMA-7.3 molecule, which is the one that is in the Blue Earth Diagnostics development pipeline right now, and will likely be another clinically available PSMA-targeted radiotracer.

So what about ProstaScint and how is all this different? Well, again, ProstaScint makes use of an antibody, the 7E11 antibody, and it is an indium-111-labeled molecule and is performed with SPECT imaging. So because of that, it has lower spatial resolution than the PET radiotracers are able to provide, and in addition to that, it has higher background levels. Furthermore, it suffers from lower sensitivity, and that is because the 7E11 antibody binds to this internal epitope on the PSMA molecule, and so cancer cells really require openings in their membranes in order for the antibody to come across and bind to it. So because of that, it is actually quite difficult for this radiotracer to bind and produce a signal which can then be viewed diagnostically.

Here are some images of ProstaScint scans. Steve is going to show us a good number of PET scans today, and if you keep these in the back of your mind, you'll see just significantly improved signal-to-noise ratio when we use the PET radiotracers, which bind to that active site in the external epitope of the PSMA molecule versus the 7E11 antibody, which, again, suffers from the fact that it is a SPECT radiotracer and binds the internal epitope. Here, we can see significant binding in the bone marrow, which is nonspecific. This patient does not actually have boney disease, but as an antibody, it binds non-specifically to the Fc receptors of cells that are contained in the bone marrow. We do see some signal in the prostate here, but there's a lot of noise.

Similarly here, in the prostate bed of this patient who underwent a radical prostatectomy, yes, we see the signal, but we see a lot of noise, and really, when we just look at the scintigraphic portion of the process and studies, we see just tremendous uptake all throughout the body. This is really why this radiotracer, which came about in the late nineties, really never took off clinically because the images really left much to be desired.

So in contrast to that, here are some images that were generated with DCFPyL, which is a PET radiotracer. We can see in comparison to a patient who was imaged with methylene diphosphonate, we can see markedly increased sensitivity here in these latter two panels, the last panel being one where we blocked out the tissue uptake and just allowed for the fair comparison of bony metastasis. This is, again, being the same patient. We see a marked increase in the sensitivity for detecting sites of disease. Here, we see a total of 87 sites of disease in this patient, whereas when they were imaged with MDP SPECT, only 12 sites were seen. Yes, with sodium fluoride PET, we see increased sensitivity relative to the SPECT radiotracer, but here we only see 39 sites of disease. So by using both PET and a significantly more sensitive radiotracer, we are able to pick up on many, many more sites of disease using a PSMA-targeted radiotracer.

Here's what images look like in clinical practice, where not only do we have that maximal intensity projection, which I was showing in the prior slides, but also we are able to see that PET radio signal overlaid on top of conventional imaging in the form of a CT scan, and these are the type of images which are more typical in clinical practice.  We see very, very little background uptake with the DCFPYL radiotracer and we see really spectacular uptake in things like pelvic lymph nodes.

So again, there are two PET radiotracers that are approved for routine clinical use in the United States. The first is the gallium-68-labeled PSMA-11 radiotracer. This radiotracer currently is available at a limited number of institutions. At first, it was approved based on a new drug application to the FDA, which was placed, combined by UCLA and UCSF. They were able to receive approval for this radiotracer, and they synthesize it at those two centers and are able to image patients with it. In contrast, DCFPyL got FDA approval by Lantheus Medical Imaging after they put forth an NDA rather, and I'll explain to you why Lantheus is able to offer this throughout the entire country, where just very selective sites can the gallium-68 radiotracer be produced.

So what are the important differences between PSMA-11 and DCFPyL? Well, so what a lot of it comes down to is not so much the ligand, both of them are urea-based small molecules, which bind to the active site of PSMA, but rather the radionuclide which is bound to them. So gallium-68 has a shorter half-life than F-18; 68 minutes versus 109 minutes. So that makes it more difficult to then be able to produce the radiotracer in a centralized location and then distribute it to different parts of the country. So this really requires that the radiotracer be produced on-site rather than being produced at a centralized location. In addition to that, gallium-68 is produced by a generator. I won't go into the details of really the difference between a generator and a cyclotron, but suffice it to say that generators allow you to produce only very small quantities of the radionuclide whereas a cyclotron allows you to produce huge quantities of it. So now we have radiotracers which are F-18-labeled, which have long half-lives, and we can make them in large quantities, making it so they can be distributed throughout the country.

There are also some reasons to believe that there is superior image quality and therefore improved sensitivity with the F-18-labeled radiotracers relative to the gallium-68 radiotracer. And that has to be the fact that F-18 has lower average positron energy than gallium-68, which leads to a shorter path length to annihilation. So if we remember back to my earlier slides, where I talked about a positron needing to be emitted so that it then causes an annihilation event, lower energy positrons, which have shorter distances to travel before they cause that annihilation event, will lead to more crisp images, because what the PET scanner is detecting is the radio signal actually closer to where the radiotracer is bound and that is a property of F-18 [inaudible 00:26:09] to gallium-68. So because of that, by and large, we see superior image quality with the F-18-labeled radiotracers.

So this is the same patient who was imaged with PSMA-11, as well as DCFPyL. The patient was imaged at the same time point, and there is the same windowing present here, and you could just be the judge for yourselves. We see a markedly improved signal-to-noise ratio with DCFPyL than we do with PSMA-11.

The two radiotracers have been compared head-to-head, albeit in small studies. This is just one paper from Dietlein and coworkers. In this study, they had 14 patients with biochemically recurrent prostate cancer with a range of PSAs from 0.4 to 50, with a median PSA of 2. These patients underwent treatment with either radiotherapy or prostatectomy and then they were imaged with both radiotracers. Greater than one lesion was detected in 10 of these 14 patients. DCFPyL identified all sites of disease which were found with PSMA-11, however, there were three patients who had more than one lesion that was detected with DCFPyL that was not detected with PSMA-11, suggesting a higher degree of sensitivity with this radiotracer. Again, this study suffers from a small sample size, but it is at least an indication that these theoretical properties of the F-18-labeled radiotracer do come to bear in clinical work. When they look at individual lesions, they see higher lesion conspicuity with the DCFPyL than they do with PSMA-11 as well. So it's not only more sites of disease, but a higher signal within those lesions.

So what are the applications for PET imaging of prostate cancer? Well, we could use these radiotracers to help guide prostate biopsy, we could use it for initial staging in patients who we believe are harboring occult metastatic disease, we could use it to direct therapy upon biochemical failure following failed local therapy, we could use it to confirm the extent of disease in patients who are felt to be oligometastatic, and we could use it to evaluate response to treatment, as well as to judge candidacy for endoradiotherapy, meaning treatment with radiotherapy that are linked to PSMA ligands. With that said, however, we do really only have two approved indications by the United States Food and Drug Administration for using DCFPyL, again, known as PYLARIFY, as well as the PSMA radiotracer.

So I'll just read the statement from the label insert, which appears identically for both of these radiotracers, and that is, that radioactive diagnostic agents indicated for PET imaging of PSMA positive lesions in prostate cancer is limited to patients: with suspected metastasis who are candidates for initial definitive therapy, so said differently, the radiotracer is approved for the pretreatment imaging of patients with high-risk prostate cancer; as well as in patients with suspected recurrence based on elevated serum PSA level, said differently, in patients with biochemical recurrence. So although we do have a myriad of potential applications, these are really the only two that we have FDA clearance for and these are really the only two then that insurance carriers will pay for.

So where do these indications come from? Well, I'll talk about DCFPyL, which is the radiotracer that Steve and I have studied extensively through our work in the OSPREY and the CONDOR trials. So the first study, known as the OSPREY trial, was part of the registration packet that the FDA received for approving PyL. The OSPREY trial had two cohorts to it. The first cohort was in patients who were newly diagnosed with prostate cancer but were found to be either high or very high risk based on NCCN guidelines. These patients were scheduled to undergo a radical prostatectomy, and prior to this, underwent PET/CT imaging. The patients then underwent the radical prostatectomy with pelvic lymph node dissection and so we had the pelvic lymph nodes with which to compare the results of the PET imaging to serve as the truth standard, to establish the sensitivity and specificity of the radiotracer.

There was also a second cohort, known as cohort B, of patients who have metastatic disease who are undergoing a biopsy of their metastatic sites and this really just served as a benchmark that the things that we were seeing with the radiotracer were indeed prostate cancer and not non-specific uptake of the radiotracer.

So this data comes from cohort B, the metastatic cohort, and what we saw is a very, very high sensitivity of the radiotracer for imaging those sites of disease. Overall, that number was north of 95%. When we looked at different compartments within the body, whether it was the bone, the lymph nodes, or the visceral soft tissue, we saw that this high degree of sensitivity was preserved. Again, sensitivity is established following the biopsy of the metastatic sites.

In preoperative staging, we looked at the pelvic lymph nodes to determine both sensitivity and specificity. Sensitivity was in the range of about 40% overall and that is because when we use surgical specimens as a truth standard, even if a lymph node has only one cancer cell in it, it was caused positive by this rigorous truth standard and as good as the DCFPyL radiotracer is, it's simply not possible to image only one cell at this time. So if we looked at positive lymph nodes which are larger than 5 mm in size, we found a sensitivity now of 60% for the radiotracer. Perhaps more importantly, the radiotracer had a remarkable specificity. When we looked at where the radiotracer was taken up relative to where sites of disease actually were, we saw specificities in the range of 85% to 97%. These bars are the [inaudible 00:32:34] result of three different independent readers who read the PET scans.

The next study that led to the registration of the DCFPyL molecule was known as the CONDOR study, and the reason this radiotracer received that second indication for imaging biochemical recurrent disease, is the result of this study. In this study, we took patients who recurred after either radical prostatectomy or radiation therapy, and then they were imaged with DCFPyL.  Then to serve as a truth standard, we had the composite truth of the recurrence being detected either by an Axumin PET scan, which was already currently approved at the time, or some form of anatomical imaging and/or biopsy. We saw very high levels of disease detection across a range of PSA values, even in patients whose PSA values were under 0.5, we saw sensitivities in the range of 36%. When we went up to PSA values as greater than 5, the sensitivity was north of 95%. So excellent sensitivity across a range of PSA values in this study.

When we compare PSMA-targeted imaging to conventional imaging, we see significantly higher values of sensitivity, as well as specificity. This was done in the proPSMA trial, which was conducted in Australia. Patients underwent both conventional imaging, as well as PET imaging, and here are some of the plots of those values with respect to any metastatic disease, pelvic nodes, or just distant metastases. In blue, we see conventional imaging; in red, we see PET imaging. We see significantly improved diagnostic performance when imaging these patients with a PET radiotracer.

PSMA PET imaging has also been compared to fluciclovine, which is FACBC or Axumin. This study was done by that same group in Australia, where they imaged patients with biochemically recurrent prostate cancer with both radiotracers, and what they found is a significantly higher degree of lesion detection with the PSMA radiotracer, relative to Axumin, and we also know that this radiotracer is more specific as well.  So PSMA PET imaging has been available at a number of centers in this country under various registry-based imaging protocols for some time. This is a study from Tom Hope's group at UCSF, where they looked to see if once you gave clinicians access to the PSMA radiotracer, in this case, PSMA-11, whether or not it would lead to changes in patient management now that this newfound sensitivity and specificity have been gained. And what they found is in more than half of all the patients who underwent this form of molecular imaging, there was actually a change in their prostate cancer management.

I should note, however, that it is unknown at this time, whether or not these changes in management led to actual clinical good for the patient, but it does, I think, stand to reason that if you are able to newly see sites of disease that you could not see previously and change the treatment paradigm for a patient, it is probably reasonable that if you avoided one treatment, for which the clinician determined it was inappropriate, that probably some good was achieved. But we really do not know, at least on a randomized basis, if patients are randomized to standard of care versus downstream management on the basis of PET scan, whether or not the patients truly benefit in terms of long-term outcomes like metastasis-free or overall survival.

There is additional thought that the PSMA-based radiotracers could service prognostic markers for prostate cancer. This is a study by van Leeuwen and coworkers, where they looked at patients who underwent PSMA-targeted imaging with the gallium radiotracer prior to radical prostatectomy with pelvic lymph node dissection. They looked at just the cohort of patients who were pathologic N1, and they looked to see what were the out outcome in patients who are PSMA-positive N1 versus PSMA-negative N1, and they found that those patients who are PSMA-positive were three times more likely to experience biochemical recurrence than those who are PSMA-negative. This is likely due to a function of the size of lymph nodes that could be detected with PSMA-targeted PET imaging. Likely patients with larger lymph nodes were detected with this imaging modality, whereas those patients who had small lymph nodes, which are below the level of detection, could not be detected with the imaging modality, neither of which could be detected with conventional imaging, but, however, once detected on PSMA-targeted PET imaging, some information became known about their prognosis.

So in conclusion, conventional imaging modalities have proven insufficient for detecting prostate cancer, particularly in patients with low PSA values, and those patients with biochemically recurrent disease. PSMA-targeted PET has improved our ability to image prostate cancer, both in terms of sensitivity and specificity, across a range of disease states. PSMA-targeted PET imaging has been shown to outperform both conventional imaging and other PET radiotracers in rigorous clinical trials. Although imaging with PSMA-targeted PET has been shown to frequently relate to changes in management, it is really unclear whether or not these changes in management are associated with improved patient outcomes, but we're only going to learn this with additional study and now we have available to us, at least here in the United States, two PSMA-targeted PET radiotracers, that being the gallium-11 radiotracer and PYLARIFY, the 18F-DCFPyL radiotracer. The PYLARIFY radiotracer is available throughout the country now because of the unique properties that F-18 can be produced in large quantities using the cyclotron and it has a long half-life allowing for it to be delivered in such a fashion, and we do have data which leads us to believe that this more widely available radiotracer actually leads to improved image quality.