Non-PSMA Based Theranostics in Prostate Cancer - Andrei Iagaru
March 7, 2020
Andrei H. Iagaru, MD, FACNM, Professor of Radiology - Nuclear Medicine, Chief, Division of Nuclear Medicine and Molecular Imaging, Director, Nuclear Medicine Residency Program, Co-Director, PET-MRI Research Program, Stanford University, Stanford, California, USA
Andrei Iagaru: Good afternoon, ladies and gentlemen. Thank you very much, Hussein and Richard, for the invitation to be here. It's an honor to be in your session about theranostics in prostate cancer. I'm sorry, I'm going to disappoint a few in the room. I'm not going to talk about PSMA, but that's all right. I've seen that there is life outside of PSMA, although I do support PSMA imaging and therapy. This is an outline of my talk. I'm going to give a brief introduction, and then I'm going to show you another target for prostate cancer, gastrin-releasing peptide receptors, and the bench to bedside trip with one of these radiopharmaceuticals RM2. I'll end with a few slides on what I think may be interesting future directions.
This paper is published every year. It is statistics of cancer in this country, so you can download it for free and see the trends for various cancers. Prostate cancer continues to hold steady as the number one in the new number of cases diagnosed in the US, and the second as far as mortality each year in our country. Definitely, this is a huge healthcare issue that we have work to do to attempt to solve. If you look on the next slide, and this is data from 2012 and 2013, which is the most recent available as far as changes from year to year, cancer is actually still a very close second to heart disease as far as killers in the United States. Again, we have a lot of work to do to try to change this. Interestingly, where there were changes and decreases in mortality from infection from heart disease, what we were able to change in cancer, a decrease of 3% in mortality is actually frankly dismal. Again, we have work to do, and this work that has to be done does not belong only in med-onco surgery. We are part of this partnership of healthcare providers who can work together and should work together to improve the care that we provide to our patients.
Now, let's talk about gastrin-releasing peptide receptors. They belong to the bombesin receptor family, which is over-expressed in a variety of major human cancers, northerlies being prostate cancer. But they also include about 80% of estrogen receptor-positive breast cancer, gastrointestinal stromal tumor, ovarian cancer, and others. This is interesting, an interesting target that has been studied for more than a decade now. At the beginning, people were trying to use agonists versus antagonists. Antagonists were introduced more recently, and in fact, right now the entire development in this area uses antagonists only. The reason for that is that there are significant side effects on agonists administered in these patients. They have been tried including by Dr. Baum's group in Bad Berka, the first in the world to do it.
There are no side effects associated with the administration of antagonists. These are the three agonists that have been tried in humans. The one on the top in patients with gliomas and GI stromal tumors. A more recent one in China at the Peking Union Medical College in patients with prostate cancer and the AMBA inpatient with various cancer, and I believe that he's the one that Dr. Baum has tried previously. This paper has been just recently published in JNM, and it did compare and it's the first one to my knowledge to directly compare an agonist and antagonist in the same patients. Why is this important? It's important because it illustrates the differences between the uptake of agonist versus antagonist, and particularly in this example where a patient with skeletal scintigraphy here on the left and on the top, you have the antagonist RM 26 and on the bottom, the agonists think they see the signal is much stronger with the antagonists compared to the agonist.
To my knowledge, this is one of the first, if not the first paper, directly comparing in the same patient and prostate cancer patients an agonist and an antagonist, very close temporally to each other. Well, they've also shown very interestingly. It's the number of lesions in a primary tumor on the left versus lymph node metastasis versus bone metastasis, and you see not only a higher signal we see antagonists but also a larger number of lesions identified with the antagonist when compared to the agonist. This conference, what we knew from preclinical studies and from some early experiences, agonists are only in prostate cancer that if we are to take this forward in humans in larger clinical trials, et cetera, we need to focus on the antagonist and not try to use the agonist. If you are to search the literature, these are the different classes of targets that have been developed for imaging of GRPR, gastrin-releasing peptide receptors as well as the compounds that have been translated clinically in humans.
The ones that are listed there are the ones that I'm going to show over the next few slides before I go over and share with you our experience at Stanford. This is the normal bar distribution. Once you inject one of these agents in humans, there's very little uptake in the head and neck as well as in the chest. The pancreas is the most important area of uptake because that's where gastrin is produced and released. There's little to none in terms of hepatobiliary clearance, and there is genital urinary clearance, you can see kidneys, ureters and bladder. There is a small amount of optics throughout the esophagus as well as in the rectum. This makes it an agent that if you are targeting things that are located in the head and neck, in the chest, as well as in the abdomen and pelvis, has very good clearance properties as well as little confounding factors as far as physiological elimination of the radiopharmaceutical.
Now, let's go through some of the agents that have been used in humans. This is one of them gallium labeled, and you will see here is from Dr. Baum's group collaborating with others at Erasmus in Rotterdam. This was a proof of principle in patients who have breast and prostate cancer and this work. I believe there were some issues with stability of this probe, so this was improved, and the next generation is called NeoBOMB1, and this is licensed by AAA for a few months now. I believe there are plans for larger studies with this agent. This was published also in JNM, and interestingly this was published in a study that paired not just the diagnostic version label of gallium-68, but also the therapy version labeled with lutetium-177. These, just like other compounds that we are looking at for prostate cancer and other cancers has the potential to be used both for diagnostic and then if you identify the target for therapy as well.
We are lucky to have Dr. Baum who coined the term theranostics here in the room and he'll give you his experience in using this in thousands if not tens of thousands of patients to date. People try to label the same compound that I'm going to talk about later on, RM2, with fluorine-18. Turns out, this didn't quite work. When you try to label the peptide with fluorine-18 which will make it easier for commercial distribution, et cetera, something changes, either the charge of the particle or some of the spatial conformation. It does change, and it makes it less appropriate for imaging prostate cancer. This is the group in Zurich where they looked at this, and they realized that this has no future for clinical translation because it behaved poorly when compared to other radiopharmaceuticals, and in this case with fluorocholine.
You know from gallium labeled bombesin agents that they should behave better than calling for the identification of prostate cancer patients. An interesting approach is to label the peptides with copper-64. Copper-64 has advantages as far as half-life. It has a half-life of about 12 hours. If you want to do those imageries, if you want to image longer or rather later after injections, this is ideal for that purpose. There's also an isotope that can be used for therapy, copper-67, so you have uniquely a pair of the same radiochemicals with two different isotopes, copper-64 for diagnostic and copper-67 for therapy. This has been put in humans as well from the same group that developed the RM2 led by professor Helmut Maecke, and we spoke on November as a PI, and they've shown that these can be successfully used in humans. In the US, the supply of copper-64 exists. However, copper-64 is expensive, and there are only a couple of sites that make it, Wisconsin and Mallinckrodt, Wash U. I think that for an hour at least, gallium-68 will be the agent of choice for labeling these kinds of peptides for prostate cancer imaging.
Let's move on and see what is the true roadmap for taking someone from bench to pre-clinical to actually translate it into humans. We're not there yet, but hopefully, get it approved for use routinely. This is the chemical structure of the RM2, the peptide, and you can see in the ring here, it's where it can be labeled with either gallium-68 or with lutetium-177. Even to make it to the first in human studies, there's a lot of work that has to be done, and these slides summarize some of that work. You want to make sure that this is safe for administration in humans and there were no observed effect levels in various areas, for example, in the central nervous system, cardiovascular, respiratory systems. These are the major areas where one would look for the toxicity of these peptides. The binding affinity has been shown to be high, particularly for prostate cancer lines that are expected to over-express gastrin-releasing peptide receptors. This uptake is specifically shown at this inhibition graph here, and this is what we want to see. We want to see is these three pharmaceuticals being taken up in areas that reflect prostate cancer or some other cancer and not in normal physiological healthy tissue. If there is, either uptake or clearance to healthy tissue, we want to make sure that, that's reproducible and you know about it ahead of time, so that you can evaluate it when you interpret the images.
There are different subtypes in different prostate cancers, and the bombesin receptor, it's overstressed as I said earlier in human prostate cancer. Also, in other types of cancer like carcinoids but also estrogen receptor-positive breast cancer. The binding to different mammalian species is still on the bottom as well as in different species that were studied such as mice and rats. This is a busy slide, but if you have a good vision on the left it's PC-3 tumor-bearing mice, and on the right, it's a different type of prostate cancer implanted tumor with different affinities for these receptors. The slide is meant to illustrate the differences in uptake between those that express the receptor and those that don't express, and it's also summarized in these graphs here when you can see that the gallium-68 RM2 in green has much higher ratios compared to different tissues when compared to FDG or fluorocholine. This, again, very encouraging early preliminary results suggesting that this peptide should be translated into humans.
Now, we're moving from cells and tissue to small animals, and these are some of the early studies that were shown that in a xenograft you can see the tumor lighting up on the projection image as pointed by the arrow, and when you block the receptor expression there is no signal which again is excellent. We want these signals to be specific. One of my colleagues looked at this together with MRI and spectroscopy and was able to demonstrate that the area that lights up with bombesin is also the area that demonstrates abnormal signals both on hyperpolarized MR, but also on histopathology. The reason why this was done because their work at that point was meant to be translated for biopsy targeting using fused pet MRI and ultrasound. Currently, MRI and ultrasound are used by urologists to biopsy prostate lesions, but we hope and we want to be able to demonstrate that you can improve your targeting for biopsy by using a specific peptide pharmaceutical, in this case, the bombesin agent.
As far as therapy preclinical work, of course, the more radiation you're able to give, the more successful that treatment is going to be. This is what's been shown here. Different dosages versus non-radioactive lutetium and they've shown that there's a nice stratification of response. Rapamycin is a radiosensitizer, so the combination of this peptide labeled with lutetium-177 together with the radiosensitizer seems to improve outcomes. This has not been done in humans yet but is definitely an area of intense interest for clinicians. What about side effects? Well, if you do it this way, even combined with rapamycin, there are not observed toxicities in the pancreas, which is that dose-limiting organ or in kidneys, which is where most of the excretion of these peptide happens. Again, this is solid preclinical data suggesting that this kind of a combination diagnostic and therapy targeting GRPR is safe as well as most likely effective.
Now, let's look at what has been done to date clinically with this combination with this pair. We have the same peptide that was shown earlier, RM2. Some of the first studies were done by the group in Turku and the group in Linz, and this is the initial biodistribution in a healthy man, where they showed that as expected from preclinical data, that target organ with the highest amount of optic is the pancreas and the clearance is through kidneys and bladder with very minimal excretions through the hepato-biliary tract. This is now moving to prostate cancer patients and comparing it with fluorocholine. This is from this study showing that the uptake in prostate cancer is higher than the one in normal tissue and interestingly and good for this compound is higher than the signal coming from benign prostate hypertrophy. That's something that's very good.
To summarize what has done at other centers, they've shown that the tumor lights up with high tumor to background tissue while BPH does not light up, so a benefit of this class of agents. This is our experience at my institution using RM2, labeled as gallium-68. This is in the initial staging of prostate cancer to date eight patients with intermediate and high risk were scheduled to and have been scanned prior to prostatectomy and nodal dissection. These are some examples of what we see. The arrow on PET and PET/CT indicates the area of abnormal uptake, and what's nice about this study is that the prostate is removed in one piece and it's put in a 3D mold that's printed from the MRI images. When you section the histology sample, you can actually very nicely correlate histology and imaging because they're based on the same MR 3D mold. It's a very nice correlation of a signal with what are annotated with ink by the pathologist.
This is another example of prostate cancer that's crossing between the left and the right side of those on imaging as well as on histopathology. This is an example indicating multiple areas of uptake in the prostate gland, and they're all shown to be correlating with prostate cancer and histopathology. A small sample, but very encouraging results and we continue to enroll in this study. We have more experience with gallium-68 RM2 at biochemical recurrence. We published data from the first 32 patients, but now we have 73 patients that have a biochemical recurrence based on rising PSA. Despite this, they have negative CT, MRI and bombesin, so inclusion criteria are to have rising PSA but negative CT, MRI, and bone scan. Despite this, we're able to find recurrent prostate cancer in about 69% of cases by PET, while MRI was able to find these in only a third of the cases and none of the cases that were found on MRI were negative on PET.
Definitely, substantial benefits in this patient population in our experience with gallium-68 RM2 for biochemical recurrence of prostate cancer. I'll go a little bit quickly through a few cases to illustrate the type of lesions that we see. This is a patient who had a biochemical recurrence since 2013, and despite multiple biopsies, MRIs, et cetera, the source of the PSA rise was not able to be found. However, I think that if you focus on these images below the prostate in the right, below the bladder in the right aspect of the prostate, you see a focal uptick and there you go on trans axial imaging, and this was biopsy-proven to be recurrent prostate cancer. Not only that, after local radiation with high-dose radiotherapy, the signal disappears and that perhaps opens another indication which is an assessment of focal therapy trying to replace the standard of care biopsy that's done currently at six months after treatment.
There's no good way to know if the treatment works short of biopsying. Biopsying the prostate comes with significant comorbidity, so perhaps we will be able to help in some patients to replace biopsy with imaging. This is another patient, same story, a rising PSA, unable to identify the source of recurrence, and in this case, recurrence was in two to three-millimeter lymph nodes in the pelvis. Now, we're in fact punting the question back to the clinicians. Before it used to be that we cannot find where the PSA is coming from. Now, we find where it's coming from, but what do they do with it? Do they find a brave surgeon who will go try to dissect the perirectal fat and find these lymph nodes? It will be difficult and probably there are other lymph nodes that we don't see. Same with radiation therapy, so perhaps there's a need for combination chemo hormonal radiation therapy in these patients.
Low PSA, no prostate cancer identified, and in this case, the recurrences in pelvic lymph nodes both on the right side as well as on the left side and more extensive additional lymphadenopathy. This is an interesting case where we're able to find focal uptake in a lung nodule and of course, you would say that this would have been diagnosed by chest CT except chest CT is not part of the standard of care for restaging patients with prostate cancer. Abdomen and pelvis CT is, but the chest is not. This was biopsy-proven to be a recurrent adenocarcinoma of prostate origin.
Probably one of the most interesting cases that I've seen to date is the next one. Let's see if this will play. If they were to guess, where is their recurrence Well, we thought it's in the liver, and we send these patients to our ER consult for biopsy and they said there's nothing to biopsy. We followed up and four months later there is something to biopsy, and this was recurrent prostate cancer on the surface of the liver capsule, the only side of recurrency in these particular patients, so probably peritoneal spread of disease but a pattern that we were not aware before, but now we're careful to not miss. Interestingly, you can have a recurrence in singular lymph nodes as well. On the left, is an FDG scan with some uptake in writing on the notes, but definitely more uptake on bombesin. On the top, FDG, on the bottom, bombesin, and this was also biopsy-proven to represent a recurrent adenocarcinoma of prostate origin. Recurrence of prostate bed is not unheard of. You see it on a top on the early images. What we do at first, a quick whole-body image and then we focused on the prostate with a standard MRI protocol of the pelvis and PET is acquired at the same time.
This is always interesting what levels of PSA correlate with positive scans and at PSA less than one, we find cancer and about 40%, we go one to two the detection rate increases to 60 plus percent. Two to five, about three quarters and more than 92% if you go to PSA more than five. Of course, if you talk to people who practice in Europe, for example, they will say that we need to change what is the definition of biochemical recurrence because we shouldn't wait until PSA is two or higher. We should find it that PSA is 0.5. I think that's a legit argument to be made, but I also think that what's important is everything that we find relevant for patient care and this, I think it's an interesting question and this graph I think would argue that not everything that we find is important.
These are graphs correlating the PSA velocity, with whether the scan was positive and negative, and while I'm sure that with RM2, you will have some false-negative scans. Perhaps a false-negatives are those underlying cancer that maybe we don't have to treat because of the PSA velocity, the speeds at which PSA increases were way higher in the population with positive scans versus the population with negative scans so that every patient with a positive PSMA or bombesin scan needs to be treated. I would argue that perhaps not, and that's a very interesting question that we need to address. My younger colleagues, residents, and fellows are working on bio-distribution atlases of these radiopharmaceuticals, and this is their work where they classified uptaking, nonsignificant and you see what organs are there, mild, moderate and high. What's important is in those lesions that were identified, the signal was pretty high. Again, a very good lesion to break ratio.
Briefly, before I end, a little bit about PSMA and GRPR, I think that most people think of one versus the other. I really do not think that there is a case. I think that we should think about it that PSMA+GRPR. They are two very different biological processes. I find it impossible to believe that the same patient along a long course of disease like in prostate cancer will always benefit from the same scan and will not benefit from multiple biological targets. I encourage you to think about this as well in terms of, there's no silver bullet in biology. What works in one patient or one stage of disease is not going to work in every other patient at all stages of disease. The more targets we have available, the better it will be for patients.
You can, of course, have patients where one agent will find more than the other or where they're both negative or they're both positive and showing the same lesions. Let's look at our data from PSMA and GRPR at the initial diagnosis. Four of these patients and all four were positive, but were we finding the same cancers? This is one example. Let's look at the transactional views. I think that here, you will see that this RM2 probably, and frankly retrospectively, you don't see much of these lesions that were seen on PSMA, but in the same prostate, you see another cancer with RM2 that you don't see with PSMA. Of course, you could argue that the prostate is coming out anyway, so why do you care? Well, I think that you would argue that if you have a more limited disease that you want to treat with a high-intensity frequency, ultrasound or radiation, you need to know about all the cancers that are present in the gland, so you don't miss one.
This is another interesting one where a lymph node is seen with both RM2 and PSMA, but I would argue that it's much easier to see the additional lymph node with RM2 than it is to see it here with PSMA where it's very close to the ureter and frankly we did not call it prospectively. What about that initial diagnosis? This is another interesting case where on the top you see the RM2 signal in one part of the gland and at the same level, the PSMA signal. It's in a completely different part of the prostate gland and we know from pathology that these were all cancers. I hope to be able to raise your interest in both targets because they're looking at different biology, and as of now, I don't know which one is more important than the other. I think that they're both very important in the biology of prostate cancer.
Let's look at the use of both at the biochemical recurrence. We published first a couple of years ago, a smaller series, but now we have 29 patients with prostate cancer and you will see here what the ratios of what were both positive when one was more positive than the other, and when one was positive and the other one is negative. About two thirds, you find the same lesions, but in the other third, there are differences. That's thought of the patient can be relevant in how we manage them and how we use these agents. This is an example of a patient who has been followed up for 40 plus months before coming to us including negative scans at other institutions, and this is one that I use routinely because it illustrates much easier seeing lymph nodes in the retroperitoneum with RM2 than with PSMA. You may argue, well you find five lymph nodes versus 10 lymph nodes, who cares perhaps, but in some cases, it does make a difference.
This is another example where, again, you see retroperitoneal lymph nodes easier with RM2 than with PSMA and similarly for a pelvic soft tissue lesion. Then you have patients where both show up around the same time, and then, of course, you have patients with an extensive disease where it doesn't really matter if you find one more, as Rod Peaks would like to call them, lumps or not on the scans. It's interesting even here to notice that with PSMA you see maybe more bone marrow disease, with RM2, you see perhaps more lymph nodes. The uptake is quite different, and this was shown for the first time here, but we're able to do a comparison now is the fluorine-18 PSMA with PYL, and this is a patient where both scans are negative. These are patients where both scans are showing the same lesion in the prostate, in the seminal vesicles, and this to my surprise is one where the bombesin was completely negative and there are multiple small bone marrow lesions on PSMA.
Future directions? This is the first patient that was treated with lutetium RM2, so quite an advanced stage of disease. I certainly hope that we'll get these patients at much earlier stages of disease when we can have an impact on longer-term survival. These are comparisons of RM2 and PSMA in these patients and this is a longer-term follow up in the same patient where the signal washes out from the pancreas, but persists in the metastatic lesions, so that's another good kinetic property of this agent.
Many things that we plan to do including biopsy guidance, monitoring, response to local therapy, the use of this in breast cancer, et cetera. If you want to read more, we probably see some review relatively recently in genomics. It's very comprehensive. I want to thank a lot of people. Helmut Mäcke is the brilliant mind before many compounds, including RM2. Wolfgang supported our exchange of experience with RM2 and the rest are the people at Piramal who took this compound from preclinical to a kit that's ready to be radiolabeled.
The Department of Defense doesn't sponsor just work. This work is sponsored by the Department of Defense here in the US, and it's an impact award for our institution, so I want to acknowledge their support. I also want to acknowledge everyone at my institution who works on this and the people are in the audience here today. Lastly, I want to dedicate this to a good friend who's no longer here. Thank you.