Theranostics - Taking Over Prostate Cancer Care? - Phillip H. Kuo

December 16, 2021

In this LUGPA CME presentation, Dr Phillip Kuo presents theranostics and its role in prostate cancer treatment. He provides a brief introduction into theranostics, a focus on PSMA, highlighting some select theranostic trials, then delves into a broader vision of what theranostics can mean, particularly to the nuclear medicine committee and PSMA. 

Phillip H. Kuo, MD, Ph.D., Professor of Medical Imaging, Medicine and Biomedical Engineering, the University of Arizona and Senior Medical Director - Invicro, A Konica Minolta Company

Read the Full Video Transcript

Phillip Kuo: Hello everyone. Thank you. And thank you to the meeting organizers for giving the opportunity for the Society of Nuclear Medicine, Molecular Imaging to speak at this great meeting. I've been given the task of this, I thought a pretty provocative title, Theranostics: Taking Over Cancer Care? And that is a question which hopefully you will perhaps be able to answer for yourself better after my talk.

If I may start by way of disclosures, I am actually a hybrid. I'm a Professor at the University of Arizona but I'm also Senior Medical Director for an imaging CRO known as Invicro and I'm a consultant speaker or I have grant funding from Amgen, Bayer, Blue Earth, Chimerix, Eisai, Fusion Pharma, GE Healthcare, Novartis and UroToday.

In way of an outline, I want to give you a very brief introduction to theranostics, which might be a new term for some of you, discuss our focus on PSMA, which of course is the most relevant one for this audience, give you the quick highlights of some select theranostic trials that I think it's important for you to know about those highlights, and then give you a little bit of a broader vision of what theranostics can mean, particularly to the nuclear medicine committee and PSMA, the range of the therapeutic molecules and mechanisms and even cancer types that may be treated with similar agents and then wrap it up for you guys in 20 minutes.

Theranostics, actually the term was first generally attributed in 1998 to a Mr. Funkhouser, he used the term in a press release, it actually had to do with cardiovascular diagnostics to describe a material that could combine diagnosis, treatment, and follow up of disease.

Theranostics, at least the way I think about it, and many people do is it is sort of a subcategory of personalized medicine. And so within personalized medicine, they can handle areas of screening, diagnosis, treatment, and follow-up. And we are going to focus on treatment here in theranostics. But at the core, in the beginning, is these biomarkers and they can be fluids, biopsies, or imaging according to this chart. But I actually like to think about it in more of an integrated diagnostic sort of way, and as you're going around the vendors, you kind of saw it as you just walk down the line.

There's genomics, pathology, and imaging and they really should come together and they will give you the information to really personalize the risk profile of your individual patients. It can act as a companion biomarker for targeted drugs for you to select the right drugs and also to assess response and even potentially an early diagnosis of recurrence, specifically for nuclear medicine imaging-based guidance. As Steve mentioned earlier, not so much in the US but in Europe, they're making great advancements in image-guided surgery. We do it for interventional procedures for radiotherapy as we're talking, not only obviously for our radiation oncology colleagues with external beam to know where to treat but also with radioligand therapy, which I will be talking about today.

Why PSMA? As Steve said, it's overexpressed in prostate cancer and tends to increase in aggressiveness but unlike what Steve was focusing on with the imaging agent, which is more at initial staging and biochemical recurrence, the main trials in radioligand targeted therapy have been in metastatic castrate-resistant prostate cancer. We are going on the other end of the spectrum. And so at the late stage, the tumor, as everyone here knows better than me, can undergo neuroendocrine differentiation. And that point, the PSMA expression can actually rapidly drop off. It's very important that we identify which patients have actually started to go off that cliff of PSMA expression.

And so, when in addition to a similar molecule like we have with PyL for the imaging agent, that same base molecule can actually be bound to a metal. And this metal is lutetium-177 in this particular example, and that is a beta emitter, which is just a highly charged electron. And that will actually cause tumor death because it can deliver enough localized radiation.

The combination of PSMA-PET, which you just heard about in this case, again, many, many patients will be totally PSMA positive, but we need to identify that subgroup of patients that actually have stopped or have decreased their expression of PSMA because we do not want to treat them. And also even in the PSMA positive patients, can we start to stratify the degree of PSMA expression? Are we delivering 1X, 2X, or 10X radiation to that tumor based on its PET scan uptake? And then we can use that to target and select the right patients for PSMA targeted therapy. This is a CME talk so I got to tell you right now, it's not FDA approved for patient selection, unlike the indications for initial staging and biochemical recurrence.

This was one of the groundbreaking phase two trials, comes out of a very successful group out of Australia and it was the 2018 Society of Nuclear Medicine image of the year. These are actually PET scan images from gallium PSMA-11. And these are the before and after PET scans, after being treated with the lutetium-177, the beta particle emitter, PSMA-617. That is a small molecule of radioligand therapy that binds the PSMA on the tumor cells. But as you can see, it does have some off-target binding. It also goes to the kidneys, the liver, the salivary glands, and the lacrimal glands. Let's just look at the top two panels on the left, the number at the very bottom was their serum PSA. As you can see that the patient's serum PSA after radioligand therapy dropped from 15 to let's just say undetectable for the sake of argument and what's in red is actually the segmented out PSMA positive disease.

And they use PET imaging here to actually follow up how good the PSMA radioligand therapy did. And you see that in the top left patient, it did extremely well. And the imaging also showed a complete response as did the serum PSA level. That was in, as you can tell, the retroperitoneal adenopathy. If you look to the one just over to the right, you can see it is also very successful many times in the treatment of bone metastases. Soft tissue and bone metastases can be hit, not surprising because remember, we're going after the PSMA target. No matter where that metastasis is, if it expresses PSMA to a high degree, we could potentially deliver that lethal dose of radiation. Interestingly, also in correspondence to this theranostic paradigm, imaging, the planning PET scan was also used to adjust the dose according to the number of sites of disease, the total disease burden, and also to control how many cycles of therapy. If the patient got a complete response, they actually held back on some of the doses. So really imaging was truly guiding the therapy in this phase two.

And for those of you who are familiar with the PSA waterfall charts, I showed this [inaudible 00:07:07] for you were, just over a majority of the patients were able to get over a 50% response in their same PSMA as in this Lu-PSMA trial.

And what did they use for patient selection? They actually used two kinds of PET scans, the PSMA-PET, which makes sense. You're looking for that biomarker that your radioligand therapy targets but they also used good old-fashioned FDG-PET. And why would you do that? Well, there's actually a general correspondence with the aggressiveness of the tumor and FDG expression. It kind of helped you to detect the PSMA negative disease because the FDG expression would kind of go up as the PSMA goes down. As you might imagine, it's harder to find negative disease because it doesn't show uptake.

And so they used that discordant high FDG expression, low PSMA expression to weed out those patients who they thought would not succeed on radioligand targeted therapy to PSMA. They actually then did another phase two trial, because that was just a single arm. In this one, they compared lutetium PSMA-617 versus cabazitaxel.  And they actually upped their game and increased the threshold, the stringency for FDG-PET and PSMA-PET.  They actually excluded even a higher percentage of patients in this trial than they did in their previous one.

And compared to cabazitaxel, there was a greater than 50% decrease in serum PSA in 66% of patients that received the radioligand therapy compared to 37% who received the cabazitaxel. So almost a doubling in the response using serum, greater than 50% drop in serum PSA.  The hazard ratio was 0.63 for progression-free survival radiographically or by PSA, not overall survival.

And also of course you should ask about the adverse events, the PSMA-617 was better tolerated than cabazitaxel in general but there were some very specific adverse events that the lutetium PSMA did not do as well. As you remember from the imaging, you saw those salivary glands, you saw the lacrimal gland uptake, those will get hit and dry mouth and dry eyes were substantially greater than in the cabazitaxel group but overall, patient-reported outcomes were better for cabazitaxel than for the lutetium PSMA.

Prognostics, I think we're already starting to get into that area. This is a very nice retrospective study that showed that prognostically if you're treating patients with radioligand therapy, so that's only for radioligand therapy, targeted PSMA, the higher your PSMA uptake on average in the whole patient's PSMA-PET scan, the better they will do. It's kind of intuitive. And we expected that and we hoped for it. And there is evidence to show that the higher the PSMA-PET uptake, radionuclide uptake, then the more corollary that you will deliver more radiation there and the patients do better. And also, if the FDG uptake is higher, patients do worse as well.

So the big phase three trial is the only phase three trial that has been completed and readout. And it's more colloquially known as the VISION study, begun by Endocyte, then taken over by Novartis. I was privileged to really design a lot of this trial.  And so two endpoints, the original endpoint was overall survival, as you might expect. And then the FDA actually, after the trial began said it would accept imaging-based progression-free survival as a primary endpoint.

I'm not going to read all this but just to emphasize, big trial, 84 sites, randomized, patients got, they tried to get a minimum of four cycles and then the fifth and sixth cycle of radioligand therapy, which is spaced out every six weeks approximately was at the discretion of the treating team.  The patient population was metastatic castrate-resistant prostate cancer, one or two prior taxane regimens. And they had to have failed an androgen receptor pathway inhibitor.

In addition to these criteria, there were also imaging criteria. All these patients got a PSMA-PET scan. It was gallium PSMA-11. And the first thing we did was you looked at the scan and you wanted to make sure there was a PSMA positive lesion somewhere on the image to make sure there was something to treat. And positivity is defined as: visually you compare that lesion to the liver and if it was greater than the liver, you considered it positive. So forget your definitions of PSMA positive for initial staging and biochemical recurrence. We're not saying something that was not PSMA negative would uptake the same as the liver is not prostate cancer. We are just not considering PSMA positive for the purposes of inclusion in the trial. So you had to have a lesion with an update greater than liver, PSMA positive to be included.

Then the exclusion criteria, we did not have an FDG-PET scan in this study and we can go into why we did that but that was carefully considered. We replaced that FDG-PET scan, which is our surrogate of aggressive prostate cancer with CT criteria. That can never be perfect but this is what we did. You only assessed lesions of sufficient size in the category for their PSMA status. Lymph nodes had to be 2.5 centimeters on short-axis or greater. So if you had a two-centimeter lymph node and it was PSMA negative, same or less than liver, it couldn't have excluded the patient. And the lymph nodes had to be pretty large, 2.5 centimeters or greater. Solid-organ metastases like liver, lung, and adrenal had to be a centimeter or greater in size to be assessed for possible exclusion of the patient.

And then bone metastases were really tricky because these patients had been treated before. You see a sclerotic bone metastasis that could be a completely healed metastasis and that's why it is negative and showed no uptake. So we didn't want to exclude all those patients. We only looked at bone metastases that had a soft tissue component of a centimeter or greater and then assessed just that soft tissue component for its PSMA uptake.

13% of patients were actually excluded by the PSMA-PET scan in combination with the CT size criteria. And we were actually aiming for about 10 to 20%. The 87% went on to get randomized two to one lutetium PSMA radioligand therapy versus standard of care. And the imaging-based progression-free survival had an impressive hazard ratio of 0.4.  The overall survival had still a statistically significant hazard ratio of 0.62, which equated to a median OS of 15.3 versus 11.3 months.

And time to first skeletal event, which I always emphasize to my patients who are getting radium because people are obviously very justifiably worried about hip fractures and such, also showed a hazard ratio of 0.5 and 11.5 versus 6.8 months, showing that you can really target the bone as well.

There were good hazard ratios in all three categories and how about AEs? There were higher rates of AEs for lutetium PSMA-617 compared to the standard of care arm. That's expected. The majority of those AEs were fatigue and dry mouth. But fortunately, most of those were grade one, two. Patients with a lot of bone mets, as you can imagine, they are going to hit the adjacent bone marrow. The lutetium-177 beta particles do not travel too far but they still will hit the adjacent bone marrow. And 12% of these adverse events led to the discontinuation of lutetium PSMA-617. So, fortunately, the vast majority of patients were able to get all the cycles despite these adverse events, and 3.6% versus 2.9% AEs led to death. So not statistically significant between the treatment and the standard of care arm.  Here are the PSA response categories for those who really wanted to see PSA data for greater than a 50% drop. It was 46% in the lutetium PSMA arm versus 7% in the standard of care arm.

So, I've shown you a lot of data. One of the things that were really interesting that we actually, when I was making up the rules kind of got a little bit of heat for in the beginning as if a patient had just one PSMA negative lesion, even if they had 80, a 100 PSMA positive lesions, the patient was excluded. And the idea was that one PSMA negative lesion could be obviously the tip of the iceberg of PSMA negative lesions and represent a signal on a much larger reservoir of PSMA negative disease that would not respond to your PSMA targeted therapy.

This was a paper that came out well after we designed the read rules that looked at a related question. How about looking quantitatively at the level of individual metastasis on PSMA-PET scans and how people do with radioligand therapy.  And just look only at panel A in the interest of time. If you only look at one lesion on the PSMA-PET scan to see how a patient will do to radioligand therapy and that one lesion is just the one with the single lowest activity on the scan, you can see that there is a big difference in survival. Supporting that even just one negative lesion is probably going to portend a worse outcome for a patient. 

And this is just a nice panel of images to show you the spectrum of disease that you can see in metastatic castrate-resistant prostate cancer from just very high expression on all the tumors to very low expression or low expression in the majority of the tumors, to just a real high heterogeneous mix.

So, quick rack, we're going to move quickly now to these last few categories.

The PSMA-11 and the PSMA-617 molecule are very similar. The treatment molecule, the imaging molecule, so that's good because you want one to mimic the other in a theranostic paradigm. These, their small molecules with rapid clearance when they're not bound to the tumor but there are also antibodies. And these are in many clinical trials already. We're going to have to think about the kinetics of and different properties of antibodies versus some small molecules for therapies. Big molecules, small molecules, we are also using different radioisotopes.  Good old-fashioned iodine-131, which we know for decades for thyroid cancer treatment is being used for therapy with a small molecule combination.  Betas, we're using lutetium-177, I-131, and also alphas. And this group knows about radium-223 chloride also known as Xofigo was the first FDA-approved alpha emitter therapy for prostate cancer. And we're using actinium-225 and thorium-227.

A quick note about alpha particles, the mass of a beta particle is 1/2000ths of an alpha particle. A nice way of remembering that is it happens to work out between a ping pong ball and a bowling ball. You can imagine which one would you rather get hit with? I'd take the ping pong ball instead of the bowling ball, but that's why we are trying these alpha particles. They can really deliver a lot of kinetic energy in a very small distance.  And in this paper that combined a few studies for a meta-analysis, it's showing already that the alpha-emitting radioligand therapies have promise.

And nonradioactive therapies, the PET scan is still, very important for determining the expression of that target on the tumors but we also are delving into the immuno-oncology space. In this phase one trial, PSMA-PET imaging was performed and then the patient was treated with a bispecific T-cell agent. On one end, there is the PSA binding moiety, on the other end, it binds to the T-cell. We can also get into the immuno-oncology space with PSMA and PSMA-PET for selecting those patients properly.

Range of cancers, we've shown that we could be successful in phase three with metastatic castrate-resistant prostate cancer. And now, as you can imagine, the same molecules are being studied earlier and earlier in the disease progression.  And I wanted to show this to this group. If you just looked at that PSMA-PET scan, you might think it's prostate cancer. This is renal cell. The idea here is that you can treat the target, not necessarily cancer. It's not specific per se for the cancer type but for the target.

And if you look at PSMA expression, it's actually expressed to a variable degree in a lot of tumors but it's predominantly expressed in the neovasculature, not actually on the cancer cells themselves. So it's a very interesting way that we could potentially be doing an angiogenesis-targeted way of therapy.

So, lots of things to consider in the landscape of radioligand therapy and how to use PET imaging to select your patients and to monitor them. There's FDG-PET and PSMA-PET that are still being used in many trials. We have alpha particles, beta particles. We have small molecules, we have antibodies. And even when you're thinking about all the different cancer types we can treat with just PSMA as all those different variations, as you can tell, we have a lot of work to do. To get back to the question, have we reached the end of the rainbow with theranostics? I still think it may be a little premature to say that theranostics will be taking over all cancer care soon but in my opinion, theranostics will soon be a part of the everyday conversation as the larger field of precision medicine.  About a decade ago we saw that transition, to where every day we are talking about that now. Thank you very much.

Neal Shore: We have got some time for some questions. I want to just, for all of you who may not be up to speed, it was a great presentation from thinking about the nuclear medicine, radiology focus but understand that at ASCO 2021 this year, the most important plenary presentation was this VISION trial, hands down. A New England paper and it's under accelerated review at the FDA. It will highly, likely get approved, lutetium-617 in the first quarter of 2022. And what is that going to mean? If you've been working and giving radiopharmaceuticals with radium-223, if you have a collaborative relationship with your nuclear medicine radiologist or your radiation oncologist, this is going to be an approved therapy, but this particular beta PSMA radioligand targeted therapy, as Dr. Kuo just showed you, it's not going to stop there.

There are multiple other conjugates, whether they are small-molecule antibodies or alpha particles. This field is exploding. And then there are trials that are ongoing right now. Many of you are probably involved with them. The VISION trial was post-chemotherapy mCRPC, heavy tumor burden, beaten up patients, hazard ratios positive for progression and survival. And now there's a whole slew of trials going on in pre-chemotherapy mCRPC and there are trials that are going on right now in metastatic castration-sensitive prostate cancer, looking at PSMA radioligand targeted therapy. It's coming. It's coming in 2022. That is why we wanted him to have this presentation. SNMMI has great literature on its websites. They are actually having a big meeting right now in New Orleans. We really appreciate that Dr. Kuo came here instead of going to his main meeting and Dr. Rowe was able to participate. But gentlemen, I know you have questions.

Gordon Brown: Dr. Kuo, it was a great presentation. Thanks very much. I think it's exciting that we're seeing even more precise delivery of therapy, based on radioligand administration. I guess based on both the TheraP trials, as well as the VISION trial, we see significant improvements in outcomes in patients who are heavily pretreated. I guess a couple of questions. One, is if these patients respond initially with CRs on their follow-up PSMA-PET scans, are we stopping therapy? Not giving them four? Or are we continuing? And two, I guess the flip side of that is how much is too much? Is there a maintenance role here for therapy downstream?

Phillip Kuo: Great question. Thank you. And so unlike the Australian trials that did, if you were such a wonderful responder that you had no evidence of disease in which they stopped.  In the VISION trial, unless there was some adverse event, you got at least the four cycles, and to go six was even recommended if you were doing well. If there was no reason to stop and a good response, didn't include a response, was not a reason to stop. You get the six. Even if your PSA is zero and your scan is completely negative, you should get those six cycles. Maintenance, those trials have not been done yet. I don't expect at all that approval will get there. We ran into that same problem with radium. Patients could have a great response and then the other patients are like, "Whoa, whoa, what now?" So that's a challenge and I cannot read Novartis's mind, but you can imagine that's probably in the potential roadmap.

Now in the VISION study, PSMA-PET was only used for patient selection. In that trial after that, no one else got any other PET imaging. You can imagine at the time if you follow the history, Novartis took the trial over from Endocyte and that would have added a great amount of expense to a trial to do the PSMA-PET follow-up. And also you didn't really, at the time there was a nervousness about okay, let's say they're not getting better. We don't want them to necessarily stop. Because a lot of these patients, stabilization of the disease is actually a victory. I do not have to tell this audience that. And so what if their PSA doesn't drop to zero, it could still be doing a good job.

Now what was also not part of the protocol but is a very nice attribute of lutetium-177, it's actually a very good gamma emitter.  So we can do great SPECT imaging. In a lot of studies, they are actually imaging the lutetium PSMA-617, which is great. It's sort of a freebie when it comes to the radiation. You gave them the therapeutic dose and you can do wonderful SPECT imaging with it as well. And that is shown that in very recent papers if you show that you're responding very well of those therapeutic doses that you're imaging off that same dose, the patients do very well as expected.

There's one little caveat that we're always worried about, but it happens, more and more evidence just says, it just happens in a small percentage of patients but that can be important. If you do PSMA imaging to assess for a response, the one loophole is PSMA negative disease could be progressing and you don't necessarily know it because you're only imaging PSMA. That is a good argument for using FDGs on your follow up but at least in the US environment, it's kind of hard to get that reimbursed, for every time as well. Sorry, for the long answer.

Ben Lowentritt: No, it's good. Just sort of a two-part. First off, is there any suggestion that this PSMA heterogeneity, I don't know if that's the right term, but that was used to exclude patients, is that different with the different diagnostic sizes? Do we know if we get less heterogeneity or more or if that's suggestive in any way between the different diagnostic ligands that are being developed that are out there? And then the second question is, as we look forward to moving this into the potential, the newly diagnosed, castration-sensitive stage, is there less heterogeneity in that population that might make it even more ideal to consider treatment in that group?

Phillip Kuo: Great question. It's a big deal, not every site has PSMA-11, and the F-18 PyL network is growing very rapidly but can you use them interchangeably? And it is asked a lot and it's been looked at in only very small studies. The short answer is, we don't know the answer but to use an absolute SUV, as you see in the read rules that we use for patients, there is nothing about that there being a required quantification. It was just the radiologist, nuclear medicine physician's visual interpretation. The SUVs are different between the two but if you do similar to what we did in the VISION read rules, which is compared to an internal reference standard deliver and instead you actually quantifiably take the SUV of a tumor and you make a ratio out of the SUV of the liver, things normalize a lot better.

I wouldn't say that there's data that you can interchange them with yet but there's no data that blatantly says you cannot. But if you are going to try that, you are almost certainly going to have to normalize it to the liver as an internal reference standard. And then as heterogeneity is a really complex term because some people mean within the whole patient.  If adenopathy next to the kidney is PSMA negative but all the bone mets are positive, there is also tumor-level heterogeneity. Or if you take out that tumor and you cut it open, and you stain it, there's a lot of heterogeneity within the PSA expression. Fortunately for us, those betas actually travel quite a few cell lengths. It could still have the potential to a PSMA positive tumor cell that beta particles can kill the adjacent PSMA negative cell. You'd prefer every single cell within a tumor to be PSMA positive but even if it has some heterogeneity within an intratumoral level, there is still potential. And it probably does account for its effectiveness.

Gordon Brown: And lastly, just from a practical standpoint, a lot of us are delivering alpha therapies within our clinics on a daily basis. Any thoughts about whether there's going to be a heavy lift or transition to deliver beta particles just on a practical level? Or do you have any insights there for us pending the approval?

Phillip Kuo: Right. Yeah. The alpha particles, as you know, for many of you that are giving it, your radiation oncologist might be giving it or you are partnered obviously with a nuclear medicine or radiology team. Alpha particles, a sheet of paper can block an alpha particle. And the amount of gamma in the radium is so trivial you don't even have to worry about it. The lutetium is different. You actually turn the patient into basically an x-ray tube because those high-energy electrons actually make x-rays and you can actually image those also if you want. And also as I said, there's actually quite a bit of gamma that comes out of these patients. You can do SPECT imaging. So they are emitting more radiation out than you will with the radiums or the alphas. You definitely do need to be more aware of the precautions.

It becomes a little bit more iodine-131, like if anyone has ever seen how we do thyroid cancer therapies because people need to keep their distance or they will be getting radiation. On the other hand, this will not require hospitalization. This can be done on an outpatient basis. It's kind of in-between. There will be more precautions required but it is an injection. It is not some long infusion. Those who may know a drug called lutathera that we use for neuroendocrine tumors requires a contemporaneous infusion of amino acids that can make people really sick. We don't need any of that. It's kind of in-between, there is a little more complexity to it but it's outpatient. It's an injection. There is a lot of this that's going to keep the barriers low to getting out there, which is what we need to do.

Neal Shore:
All right. Fantastic. Dr. Kuo, thank you so much. Doctors Brown and Lowentritt, thank you very much.