PSMA-directed Therapy - Scott Tagawa and Michael Morris

June 30, 2020

PSMA Targeted Therapies in Progressive Metastatic Prostate Cancer

Module 4 
10 minutes: Treatment considerations following 177Lu-PSMA - Presented by Scott Tagawa, MD, MS, FACP

17 minutes: PSMA Imaging: A Necessary Gateway for PSMA-Directed Therapy? - Presented by Michael Morris, MD

24 minutes: Discussion Moderators Phillip Koo, MD, Alicia Morgans, MD, MPH, and Neal Shore, MD, FACS. 


Scott T. Tagawa, MD, MS, FACP is a Professor of Medicine and Professor of Medicine in Urology at Weill Cornell Medicine, and an Attending Physician at NewYork-Presbyterian – Weill Cornell Medical Center.  He serves as Associate Program Director for the Fellowship Training Program, Medical Director of the Meyer Cancer Center Cancer Clinical Trials Office, Associate Director of Clinical Research for the Division of Hematology and Medical Oncology, and Chair of the Feasibility Committee for cancer research. Dr. Tagawa is the WCM principal investigator for the Alliance for Clinical Trials in Oncology (formerly CALGB), serving on the Board of Directors and as a funded member of the Genitourinary Committee. He is currently a member of the Clinical Practice Guidelines Committee of the American Society of Clinical Oncology (ASCO). 

Michael J. Morris, MD Medical Oncologist, Clinical Director, Genitourinary Medical Oncology Service & Prostate Cancer Section Head, Division of Solid Tumor Oncology, Memorial Sloan Kettering Cancer Center

Supported by: Advanced Accelerator Applications (AAA), a Novartis company

Phillip J. Koo, MD, FACS Division Chief of Diagnostic Imaging at the Banner MD Anderson Cancer Center in Arizona.

Neal Shore, MD, FACS, is the Medical Director of the Carolina Urologic Research Center. He practices with Atlantic Urology Clinics in Myrtle Beach, South Carolina

Alicia Morgans, MD, MPH Associate Professor of Medicine in the Division of Hematology/Oncology at the Northwestern University Feinberg School of Medicine in Chicago, Illinois.

Read the Full Video Transcript

Lecture Presentation:

Alicia Morgans
: Hi, my name is Alicia Morgans and I'm a GU medical oncologist and Associate Professor of Medicine at Northwestern University. I'm so delighted to have you here today to talk about precision medicine, PSMA-targeted therapies, and progressive metastatic prostate cancer, in educational forum. Today we're going to talk to two really phenomenal medical oncologists who have helped to pioneer the advance of PSMA technologies, both in terms of imaging, but also in terms of therapeutics. And we'll start with Dr. Scott Tagawa who's a Professor of Medicine and Urology and the medical director of the GU Oncology Research Program at Weill Cornell. He'll be speaking about treatment considerations following lutetium-177 PSMA. Thank you so much. Scott.

Scott Tagawa:  Thank you very much for the invitation to talk about one of my favorite topics which is PSMA-targeted radionuclide therapy. I'm going to give a little bit of background about what is PSMA-TRT. I have a quick comment about efficacy but really focused the talk on once someone's been treated, what to expect.

So PSMA-TRT is prostate-specific membrane antigen-targeted radionuclide therapy which involves the administration of systemic radiation, so intravenous administration, which winds up in PSMA-positive cells. So the radionuclide component, I'm really going to focus on the beta-emitter lutetium-177, but there's also alpha and other emitters. And we have to consider the gamma emission of any particle and the differences of the PSMA-targeting vehicle.

So my one comment about efficacy is- what you can see at the bottom is we know that there are, from the available studies, we know that there's frequent PSA-declined ranges. And that does look like it happens in a majority of selected patients. Most often does happen relatively early so after the first cycle, although more happens with subsequent cycles. And it appears that patient report outcomes are favorable also in terms of the symptomatic improvement tends to happen relatively early, but there are some late responders which is something to keep in mind as we think about, "Oh, the first cycle it didn't really work; should we give the second cycle?"

To go back to the background, there's large carriers and small carriers. So the antibodies are large carriers; they have a long circulation time and because of the size they only really have access to PSMA-positive cells that are in the prostate or in tumors. But because of the long circulation time, the bone marrow and the reticuloendothelial system delivers its dose. Small molecules, also termed ligands, inhibitors, or peptides, are small by definition and they are in the circulation only for a number of hours, rapidly diffusing to PSMA-positive areas that also include quote "on-target but-off" tumors, so kidney, salivary, lacrimal glands, and small intestines. And then they rapidly diffuse and leave via the urine.

So, based on that, we would expect differences in toxicity. And you can see here we compared 130 patients approximately with Lutetium-J591, which is a large antibody, versus 50 patients with Lutetium-PSMA-617, a small molecule, and, as expected, more myelosuppression, at least with the beta-emitter, with the antibodies' circulation through the marrow, and more toxicity in on-target but-off tumor organs such as salivary glands and the small intestine.

I'm really going to talk about Lutetium-PSMA small molecules, most specifically Lutetium-PSMA-617. And anyone who searches in PubMed will find a number of publications, mostly retrospective at least initially, wherein that the Germans, a group of eight centers got together and published on 145 patients. And you can see their toxicity profiles here just magically by moving into three prospective trials, really the first three to be presented and/or published, you can see just the numbers changed dramatically just by following the patients carefully. And documented, you can see that dry mouth, nausea, and fatigue happen in either the majority or significant minority, but not a lot of high-grade myelosuppression unlike the antibodies.

There's one randomized trial that has been presented recently at ASCO and you can see here just differences between chemotherapy and Lutetium-PSMA-617, the small molecule. Again, not so much in the way of myelosuppression at least in terms of neutropenia. There was more thrombocytopenia in this series, but dry mouth, dry eye, and nausea, you can kind of see at the bottom.

So just walking through, for someone who is seeing these patients and counseling patients, what do we have to look for? I'm just going to go through these different ones. Xerostomia is one of the most common. One of the nice things is, even though it's common, it generally is low-grade and transient, lasting for days or weeks. We've investigated in a number of different preventative strategies and we haven't really found one other than using something that doesn't hit the salivary glands like an antibody. There are some recommendations for instance, in the Phase III Vision trial, sodium bi-carb rinses, cooling is unclear if it works. But, really, it's in the supportive care realm and counseling and luckily rarely permanent or high-grade.

Dry eye is pretty uncommonly reported. Actually, the highest report I ever saw was in the therapy study. Usually it's in the single digits. But, very similarly, low-grade and transient, so really I'd put it in a supportive care realm.

Renal failure is probably one of the most feared toxicities, but luckily it's rarely reported, I think because the potency of beta-emitter is not enough to cause damage, but also maybe because we're not following patients long enough. So I would just say it's something that we want to watch out for on a long-term basis. We believe that flushing the system is probably better, so, if not giving saline intravenously, then recommendation for oral hydration at least for several days and frequent voiding is part of that.

Nausea is common but also typically low-grade. So the VISION study really said that patients should get a [inaudible 00:07:06] prior to the injections plus corticosteroids for three days. We haven't done that. Generally, it's low-grade and nothing has been needed, but, like with other causes of nausea, antiemetics typically do work to some degree and vomiting is relatively rare.

Myelosuppression is reasonably common if you look at a grade 1 or grade 2, but is pretty rare if you look at grade 3 or grade 4. Looking at meters and the three- to four-week range, depending on the dose and schedule of what's used, generally speaking, only monitoring is needed. [inaudible 00:07:44]  renal toxicity is one of the things that we have in the back of our minds is what could happen in the long run, and, frankly, we just don't know. We do have a decade of experience with the antibody that's published and I do expect that to be more toxic to the bone marrow and there were no reports of MDS or leukemia or the inability to give chemotherapy following. But unfortunately these patients don't necessarily live for ten years.

Fatigue is a common non-specific side effect I think from just the radiation that's in the body. It typically is low-grade, happens relatively quickly, and most patients get better relatively quickly. I typically haven't had to do much, but I have found in those that have issues low-dose corticosteroids can be helpful or other types of supportive care. If it's severe or persistent I really look for other causes because it's not very common to be high-grade.

Pain flare has been reported. It's maybe an aberration in our series; it was around 80%. If it happens, it happens, in my experience, in patients that had pre-existing pain and happens within hours to days and then gets better over time. It's pretty rare in my experience that it happens and I would not call it a flare if it is just a lack of effect of the drug and it's progressive cancer.

So one of the things that is different about this type of therapy as opposed to most of the other therapies that we use in oncology is that there are precautions for others mostly related to the gamma emission component. The extent and duration of the precautions have to do with the amount of what's delivered. So there's on-site parameters, generally speaking, so it's locally and regionally governed. Generally speaking what we give in the United States, they can go home within hours of their injection. We do recommend oral hydration, frequent voiding. Most of the safety officers will recommend double flushing; whether it's important or not is unclear, but just go ahead and be safe.

Minimalize prolonged, very close contact. [inaudible 00:10:00] contact, ignoring the pandemic right now, is allowed- hugging, etc., holding babies for a short time but just not for a prolonged period of time. Generally our radiation safety officer advises separate sleeping, at least temporarily. It's based on the dose and what's measured after the treatment. It usually equates to three to four days. Honestly, none of my patients will follow it, but it's a recommendation that we advise laundering soiled clothes separately if that happens. Another thing just to kind of keep in mind is that they may set off security alarms. It happens sometimes with diagnostics, but they should just be counseled. And generally we give them a letter to carry with them or they call the center, say, "This was actually a therapeutic. Don't arrest them."

So, in summary, PSMA-targeted radionuclide therapy does involve a systemic delivery of radionuclides, so got to think about that in terms of the toxicity. Which toxicities are seen to which degree are in part related to the [inaudible 00:11:03] and bio-excretion of the carrier. I'd say, on average for a lutetium small molecule, look for xerostomia and nausea most commonly. And then it is a modality of therapy that does involve some precautions for others, especially related to the gamma emissions.

Alicia Morgans: Thank you for such a fantastic presentation, Scott. We are now going to move on to Dr. Michael Morris who is the prostate cancer section head and a GU medical oncologist at Memorial Sloan Kettering Cancer Center. Michael will be talking with us today about PSA imaging, a necessary gateway for PSMA-directed therapy. Thank you so much for your time, Michael. Please proceed.

Michael Morris: Thank you so much. It's really a pleasure to have this opportunity to talk about this plaguing question that has really come up in the field as we develop PSMA-directed therapy moving forward. And that is whether patients will require a PSMA image that needs to be positive in order to qualify for therapy. In order to really answer that question, I think it bears reminding that, really, the fundamental crux of that question is: is PSMA imaging a predictive biomarker for clinical benefit for patients who receive PSMA-directed therapy?

And just as a reminder for everybody about what really qualifies a biomarker as a predictive biomarker in contradistinction to a prognostic biomarker where, in essence, the incremental treatment benefit is the same if the patient has that biomarker versus if they don't have that biomarker if they receive treatment. A predictive biomarker, essentially the clinical benefit is passed through the presence of that biomarker. So a patient with that biomarker, in this case, a positive PSMA image, would be anticipated to have a much more significant treatment advantage than a patient who doesn't have that biomarker, or a patient who has the biomarker but doesn't receive drug on a trial, or a patient who of course doesn't have the biomarker and doesn't receive the drug.

So, as you can see, it's a specific kind of a trial that would prove this point one way or the other and that kind of a trial hasn't been done. And so we don't really know the answer to that question. But at least on the basis of the trials that have been performed, PSMA-based imaging traditionally has been a component of the eligibility criteria of these clinical trials. So let's go at least through some of the potential uses of imaging in terms of trying to select for patients who might benefit. Specifically if they have a positive PSMA image when receiving PSMA-directed therapy.

So, for one thing, PSMA-based imaging can rule patients in for treatment; that is, if they have a positive PSMA image, then it might be appropriate for them very specifically to have PSMA-directed therapy. And that goes back to the old adage in nuclear medicine that if you can see it then you can treat it. By contrast, and this has been used to some degree in existing trials, PSMA-based imaging in combination with other imaging modalities have been used to negatively select for patients. That is, if PSMA imaging is negative, or if FDG imaging is positive and PSMA imaging is negative, then those criteria have been used in order to exclude patients from clinical trials on the thought that if they don't have the target then they won't benefit. And then, finally, imaging can be used to make decisions about whether patients should continue to receive treatment or should discontinue treatment based on data that might suggest that we know what a good scan looks like versus a poor scan looks like in predicting whether patients will clinically benefit by continued treatment versus discontinuing treatment if it looks like they're progressing on it.

And then finally PSMA-based imaging in the age of theranostics can be used to determine dose based on the dose symmetry; that is, the delivered dose of radiation therapy to the tumor versus the normal organs in order to optimize the dose. And only treat patients who look like they will purely benefit with a minimum of toxicity based on the amount of radiation that one anticipates based on these images to be delivered to tumor versus normal organs.

Now as you can see in this chart which I've put up the eligibility criteria for a number of clinical trials. These are the larger prospective trials in which we see the use of PSMA-based imaging. So if we go back to Mike Hofman's original two studies, the first of those two studies used an inclusion criteria for PSMA-based imaging where any lesion that had an SUVmax that was greater than 1.5 of the SUV uptake of the liver brought the patient into the study. But if the patient had lesions that were evident on FDG imaging that were FDG-positive but PSMA-negative, those patients were excluded from that trial. So he used a combination of PSMA and FDG avidity to either bring patients in or exclude patients.

By the same token but using different criteria in the TheraP study which he recently presented at the virtual ASCO meeting this year, in that trial, bringing the patients in or allowing the patients in would be a PSMA image in which an SUVmax of greater than 20 were seen at any site, or any measurable site which had an SUVmax of greater than ten. But excluding patients would be those who had FDG positivity and PSMA negativity at sites of disease.

The VISION trial, which hasn't been published yet but which has completed accrual and which is the registration study for PSMA-617 that's lutetium-177 label, uses different criteria still in which patients are included if they have a PSMA-positive lesion using the liver as the base line without specific quantitative SUV cut-offs. No FDG imaging was included as part of VISION. But excluding patients from the trial would be PSMA-negative lesions that either measured greater than two and a half sonometers in nodes, greater than one sonometer in lung and liver, or bony disease with a soft tissue component that were greater than one sonometer. The Alliance trial which will look at PSMA-617, that's Lutetium radio-labeled in castration sensitive metastatic prostate cancer, which Scott will be the co-PI of, in that trial there will be no PSMA imaging and no FDG imaging for eligibility criteria.

And so, as you can see from this grid, there is not a consistent method by which studies are incorporating these scans into their eligibility criteria. What will be interesting to see of course is how these overlap. So if you apply these different definitions to the data sets from these trials, would you arrive at the same group of patients, or would you not? And would these criteria predict for a better clinical response, or not? In order to really test that of course you have to include patients who do not use PSMA imaging in order to assign them to treatment or not. And none of these trials do that. So these trials will really not answer in any definitive way whether PSMA-based imaging is a predictive biomarker.

I just want to show this as an example of Mike Hofman's methodology for both including and excluding patients on the basis of a combination of PSMA and FDG imaging. You can see here this is from his presentation of the TheraP study at ASCO. But you can see that on the far right here if a patient had PSMA-avid disease but that was FDG-negative, well, that patient would be eligible. And, at the same token, if he had concordant disease with PSMA-positivity and FDG-positivity, well, that patient would be eligible. But if they mismatched such that there was discordant uptake in favor of FDG and not PSMA, such as is seen here on far left, that patient would be excluded on the presumption that that patient has PSMA poorly expressing disease and on the presumption that that patient wouldn't be a good responder. And again I have to say that that's a presumption because these patients were excluded from the study, so it's really not testing the biomarker; it's just making a presumption in hopes of enriching the patient population for responders.

I had mentioned before that serial PSMA imaging could be used as an indicator of either progression or response. And certainly we've all seen these images in which the scans really clear quite nicely with a variety of PSMA-targeted radio therapeutics had in conjunction with PSA declines. So these certainly strongly suggest that there is a treatment effect that is a favorable anti-cancer effect, and, if not against cancer in general, at least against PSMA-expressing tissue. But it's important for us to recognize that this doesn't mean we really know what a response looks like or progression looks like on PSMA scan, nor do we know what degree of change in either direction should indicate that a patient should continue therapy or should discontinue therapy.

We don't have a standardized definition of what progression looks like, whether that's new lesions, an expanding degree of uptake in terms of distribution of disease, a rising SUV, and, if a rising SUV, then at what threshold. And we don't know whether a figure like this, for example, actually is associated with an increase in survival. We do not have the quantitative data to really consider this to be a response. I would see this as a pharmacodynamic effect, but not a biomarker for clinical benefit until we have the correlation between some measurable quantitative metric on this, whether that's SUV, or lesional size, or distribution, or the total SUV uptake in the patient- some actual quantitative measure of the PET scan that is associated with some quantitative measure of treatment clinical benefit. And again those trials aren't done and it's an important question that we still have to answer in the field as to really what clinical benefit is reflected by a series of images like this or, in the other direction, of progressive disease.

And then the last role that I'd like to at least raise in terms of the role of imaging in selecting patients is the idea of truly establishing the dose in relationship to selection of patients. So the whole idea of theragnostic is that you can see where the tracer or ultimately where the therapy will be directed to. And, again, this is from Mike Hofman's analysis of his first trial in which he showed after the fact that the patients who had more of a radiation dose delivered to their tumor had more of a PSA response. And, again, this is important in understanding who's going to respond or not. This doesn't reflect who will clinically benefit or not, but it does suggest that there may be a role for using pre-treatment PSMA imaging to establish how much radiation is going to be delivered to an individual patient and, therefore, how much benefit that they may confer by treatment. And you could decide then on that basis who should receive treatment and who shouldn't.

So, in summary, I just want to point out some of the pros and cons, at least potentially, of using imaging in the context of deciding who will get treated with PSMA-directed therapy and who won't. The pros are, of course, as I've just discussed that you might maximize the likelihood of actually treating responders and, whether that's initiating treatment or continuing treatment, imaging might be an assistive technology that will help you make those decisions. By the same token, imaging might minimize the exposure of patients who are ultimately destined not to respond and so you would spare those patients unnecessary therapy. Imaging might help you make decisions as to who should continue therapy, or, if patients aren't responding, how to define "not responding" and then subsequently to discontinue therapy. And then, finally, imaging might be used to personalize the dose that patients received in order to maximize benefit and minimize side effects. And I say "might" because none of these points have actually been shown prospectively with clinical benefit as its associated treatment effect.

Now the potential cons are real as well. So, first of all, having at least one PET scan, much less serial PET scans, much less serial PET scans with multiple modalities, significantly increases the cost of care. And, as we know, those costs aren't necessarily equally borne by society and so those increased costs will increase health disparities as well. Not every society's health care system can afford those technologies and not every healthcare system has even the PET imaging available. It will increase the treatment burden for the patient because then you're not only talking about the burden of treatment but the burden of either one modality of imaging or multiple modalities of imaging.

And then, I think, intellectually, as we think about these drugs going forward, we do, in the back of our minds, have to question really whether you do need to see disease to treat it because we understand from many other solid tumors, as well as prostate cancer, that adjuvant therapy, which by definition can't be imaged, is of value. And if we and many of us, myself included, believe that there is a value to at least exploring this imaging methodology for micro metastatic disease, so if you really believe that imaging is necessary, well, that really does undercut the concept of adjuvant therapy as well as non-metastatic disease, whether that's non-metastatic CSPC or non-metastatic CRPC.

And, again, I do believe personally that there is a role for testing these drugs in disease that can't be imaged. And until one proves the fact that imaging does confer real benefit and the fact that a negative image implies no benefit, I think that those concepts should be developed. And I think that we have to not stand on both sides of the line saying, "Yes, I believe that imaging is necessary, and, yes, I believe that patients who have non-imageable disease might potentially benefit."

The other thing is that all imaging modalities have a threshold of detectability. And there will be a micro metastatic component of this disease even with PSMA imaging, which is much more sensitive than standard imaging. And, unless we are saying intellectually that we believe that that disease can't be targeted and targeting that disease won't be of benefit, I think you have to, to some degree, have real equipoise on this issue of whether imaging is necessary, recognizing that even PSMA imaging just simply won't detect some disease but ultimately will need to be treated if you're working towards patients ultimately being cured with treatment as opposed to just putting the disease to bed for a while and having the patient recrudesce.

I think that that's my last slide and I think we can have a nice active discussion of many of the points that Dr. Tagawa and I have made and I open up the floor today.

Faculty Discussion:

Alicia Morgans: Great. Thank you both so much for these really complimentary talks about how we follow patients who are receiving treatment and how we choose those patients in the first place because these are both really critical parts of how we use these kinds of agents should they become approved. So I have many questions and I'm dying to ask, but I will let Dr. Koo, who is our nuclear medicine physician, start off by asking these med oncs  what you are thinking about when you see all this data as someone who participates in the care of these patients?

Philip Koo: Those were amazing talks and even at the moment I have a thousand things running through my brain and like your talk was so thought-provoking. And, as an imager, often times we think seeing is believing or the picture's worth a thousand words, but it could be 950 wrong words. And I think right now in the discussion about the VISION trial or Lutetium-177-PSMA, there's a lot of discussion about FDG PSMA. We're kind of simplifying it a little too much and jumping to conclusions, so I really appreciate you bringing to light this fact that we need to slow down and investigate this sort of in a step-wise, the scientific method fashion. So I thank you very much for that.

Scott, your talk was equally amazing and thought-provoking. I love the fact that we have more data and I think we are going to continue to learn how best to manage these patients as they go through the various therapies. As a nuclear medicine physician, it's interesting because we are used to treating patients and sending them back to the medical oncologist, or the endocrinologist, or what-not. When radium was first introduced, it was the first time we actually had to see a patient longitudinally. And it scared the bejeezus out of us because it's not what we do. We're diagnosticians and people that were focused on imaging, so this kind of put us into a different sphere.

Now, with Lutetium-177 DOTA-TATE and then eventually Lutetium-177-PSMA, this idea of nuclear medicine being involved becomes much more real. What are your thoughts and some advice that you would have on how we can best maximize those partnerships with nuclear medicine/medical oncology, especially as this field continues to grow?

Scott Tagawa: Just one comment that I'll make is I think that everyone here probably, when we do this, the relationship is primarily with nuclear medicine. But I think in the real world there'll be some that it's radiation oncology so just to kind of point that out there. What I'd always thought and what I think is true with this is the multi-discipline team is always going to be best for the patient. Where there may be times where one sub-specialty might be more important or leading, others more than others, but I think we always have to think about it.

Radium, I think, is a good example. That there's a lot of patients with metastatic castration resistant prostate cancer with predominant bone mets that don't ever get that drug even though we know it has an over all survival advantage and patient report outcome advantage. Part of it's because we don't think about it because it's not our specialty, but if we add multi-disciplinary teams when there is rad onc and nuc med management as part of it. And part of it might be diagnostic, part of it therapeutic, and they'll say, "Hey, what about this?" And it may not be the right thing for that patient, but if we don't think about it, it's never going to be there. Just like PARP inhibitors if we don't ever check genomics, we're not going to have any patients who are eligible.

Michael Morris: I'd have to re-iterate what Scott said that I think it's a really positive thing that we are working across disciplines to take care of patients. Everyone brings something to the table in that circumstance and I find that actually that will be true of drug development as well that you come at the development of these drugs from very different places and each field brings expertise in both the science of that, the clinical practices of drug development, and the regulatory sciences as well. So the more that nuclear medicine, radiation, and medical oncology integrate each other in both the clinical care of these patients receiving these drugs and the clinical trial design, better off that the patients are, the better off that the drug development process will be. And ultimately we'll learn a lot from each other that will only be to the patients' benefit.

Alicia Morgans: I would definitely agree and thank you for raising that, Phil. I think it's an important concept to really highlight as we try to work together in a field that we've developed a little bit with radium at least in the prostate cancer space. But I think we're going to have to continue to develop and strengthen those partnerships. And of course urologists are going to be part of the partnership and I'm wondering, Dr. Shore, what your thoughts are on all of this really fascinating data from a urologist's perspective.

Neal Shore: Thank you, Alicia. And I echo Phil's comments. Two excellent presentations. Let me ask you both a similar question and then a specific question to your talks. Michael, you gave a really great, a really intellectual overview of how we need to be thinking about trial development amongst a lot of conflicting study designs and really around the imaging. And we talk about biomarkers, blood-based, urine-based, tissue-based, but clearly imaging, what would be the role?

So the common question for the both of you is parlance and verbiage. Scott, you talked about targeted radionuclide therapy, or TRT- or maybe that might not be the best acronym or expression- as opposed to theranostics. So I'm curious for both of you to weigh in on targeted radionuclide therapy versus theranostics and what your thoughts are on that. That's the common question with the two of you.

The individual one is, Mike, I know you've done a tremendous amount of work with PyL and the CONDOR trial and the concept around this localization rate or confidence of localization rate- yet another acronym. If you could weigh in on that.

And then, Scott, if you could talk about... We heard that and positively we don't see a lot of myelosuppression with repeat treatments, repeat cycles, but do the salivary gland and/or renal toxicity, do those risks increase with repeat administrations? And I ask these for the urologists listening who are probably very unfamiliar with a lot this trial landscape, but yet it's going to be upon us before we know it. And I think it's fantastic for all the reasons that were stated.

Michael Morris: So I think that to sort of tackle your first point first in terms of how we approach the imaging I think that there may be a difference between the medical oncology world and the nuclear medicine world. Just in terms of what assumptions we make in terms of biomarkers. So, in the medical oncology world, we assume upfront that we have no idea if a biomarker has any predicable validity. So that we would say "I don't know that PDL-1 expression is necessary for a response or clinical benefit, so I will test that." Whereas I think that because of the deep rooting in imaging in the nuclear medicine community there is a presumption that, if I have this tool that can illuminate disease, wouldn't I want to use that tool to the best of its ability to select patients who will benefit?

So I think that there is somewhat of an inclination to go in with a presumption of positivity. But, on the other hand, I think that we're going to have to actually stress that assumption because the social costs are so high. Plus I don't think that we should make a presumption also that- and we'll need to test it- what degree of PSMA-positivity is necessary, if any at all, in order to benefit from treatment? Because you don't ultimately want to deny patients who would otherwise benefit. You also don't want to over treat patients who wouldn't.

So even if we were to go forward right now saying, "Hey, these trials do use PSMA-positivity as eligibility criteria," at some point, we're going to have to back our way into the question and say, "Hey, let's say we do treat patients who don't meet those criteria; do they enjoy some benefit?" At least get that question answered.

In terms of the term "theranostics," it is a combination of therapy and diagnostics. And I think that, as terms go, it has value, but I wouldn't overstep that to say that we do know that you have to have a positive diagnostic in order to have a positive treatment effect. But, certainly for enriching the patient population most likely to respond, having the positive image certainly is intuitive that that's the patient population you would focus on.

Oh, and you had asked about CLR and I think that that really does speak to the crux of the issue that every imaging modality has some threshold below which it doesn't function well. And CLR is a term that stands for Correct Localization Rate that the FDA had made to establish basically what positive predictive value would be for imaging studies. It's a combination, of course, of positive predictive value and lesional localization to make sure that your positives at least match up in the region if not at a lesional level. But the CLR of PyL was excellent at around 85%, but that does leave room for false positives and CLR doesn't speak to negatives because it's positive predictive value. So everything in that calculation is only based on positive lesions, not negative lesions. And I think that it just speaks to the fact that PSMA imaging has its limits, just as any imaging modality does. And its relationship to treatment, of course, needs to be defined as opposed to assumed.

Alicia Morgans: Great. Thank you and, Dr. Tagawa, what are your thoughts in terms of the cumulative effect potentially of some of the complications that patients can face or experience when they're receiving lutetium treatment?

Scott Tagawa: The basis for the therapy is really radiation, and the data that we have is really from external beam, but it basically shows that every organ has a certain dose limit. And that dose limit does nothing in that 100% of people are going to have or that 100% of those organs are going to have significant toxicity, but it is related to dose. It's highly related to both dose and schedule. So I would expect that organs that are exposed to higher doses for a longer period of time are going to have more toxicity. Some organs kind of naturally protect themselves, like the kidney. So I do think there is a risk for longer term, at least renal insufficiency, whether it's bad, we don't at all. It's all kind of risk benefits, but the bone marrow and the kidneys are what I'm most interested in and cautious about in terms of long term follow-up.

Neal asked about the salivary glands as well. So, in my relatively limited experience, what I've seen is that it's not so different than what I might expect with chemotherapy and neutropenia in that it's fairly similar per cycle but not so cumulative. That's with limited dosing. So whether someone gets a course this year and then a few years later get a course of something else, maybe. That's quite possible.

Shifting gears a little bit to a slightly different treatment. With the antibody using an alpha, we had six patients have some xerostomia. Only one never was exposed to Lutetium-PSMA, so whether it was pre-existing from the prior therapy, I think so, in those five of six, and it just got tipped over. But that little bit of non-specific radiation or was us asking the questions a lot in a non-randomized route, I don't know. But I do think that the salivary glands are related to intensity of radiation as well as the cumulative effect.

Alicia Morgans: Well thank you for explaining that because I think that'll be really very important clinically as people try to use these agents. I have the opportunity of asking each of you one question before we wrap up. So I'll state both questions and then whoever wishes to go first can go first.

Michael, I'd love to hear your thoughts flushed out a little more on the idea of using Lutetium in an adjuvant setting where we do not have a positive scan. Actually that's an area where I think it could be really helpful too and have even thought about certain trial designs that I might think about in that area. So I think that that's really fascinating so I'd love to hear that.

And then, Dr. Tagawa, I'd love to hear your thoughts on an interesting part of the TheraP trial that wasn't really discussed much as far as the presentation that was made because, of course, time constraints. But it was really interesting to me that it seemed that patients who seemed to have a really robust response to their Lutetium on a follow-up PSMA scan actually had deferred further Lutetium until they started to have expression. And perhaps I misunderstood the way that trial design appeared in the way that I heard that conversation, but I'd love to hear your thoughts on that aspect of the trial and whether that is something that makes sense clinically. Certainly not studied in detail yet, but what your thoughts are. So whoever would like to go first.

Scott Tagawa:  I guess I'll go first with that. So what was built in that study design is essentially Peter Mac's and Australia's experience with that drug and, in my viewpoint, coming from a nuclear medicine perspective. So what they used were SPECT scans. So post-treatment- so that treatment's already in, you can't take it out- they looked at PSMA uptake basically. And if a patient ever got to the point that he looked like he was having a very good response and there was negative imaging at that time point, they held the next dose until there was some sort of progression. Presumably it was mostly PSA.

And those were the only patients that got a subsequent PET to see if there was something that was eligible. So the minority of patients that essentially [inaudible 00:47:21] not by the way that we have defined it but in that way with- [inaudible 00:47:26] I think it was seven out of the 99 or 100 patients met those criteria. And my recollection is six out of seven got subsequent therapy. And in John Miles's publication that had 50 patients so that in the initial 30, there was a lot of that data that was in there as well for patients there.

We don't know it's different for the medical oncology perspective the trial that Dr. Morris and I designed that you mentioned before, with others, it was just a different perspective. Medical oncology, we treat until we can't see anything more. We give them more in terms of consolidation versus they wanted to back off. We don't know which one is correct. One theory- I'll cut myself off in a second; sorry- one theory that it's maybe more harmful with less of an adjuvant [inaudible 00:48:13] so there's more to go to other areas. I haven't seen that I think because the small molecules are excreted so quickly from the urine, but it is a thought and that is the rationale. We do not know if that is the correct approach; it's just the approach that's been used in a couple of prospective trials.

Alicia Morgans: Great. Thank you. And, Michael, just to hear your thoughts on the adjuvant study you would design or what your thoughts are on the use of Lutetium in this setting of not being able to see that biomarker on a scan.

Michael Morris: I think that, ideally probably for a micro metastatic setting, whether that's the adjuvant patient with high risk disease who's undergone now primary therapy, or whether it's the rising PSA circumstance, that biochemical relapse, where the PSA is the definitive indicator of relapse, probably the radioligand that you would want to use is an alpha emitter as opposed to a beta emitter. Because of the higher energy that's delivered at a lesser depth of penetration.

I think that if you look at most of our failed attempts to cure patients with adjuvant therapy or in the rising PSA context, it's because of disease heterogeneity and heterogeneity of response and resistance and persistence after the application of either hormonal therapy or hormones and chemotherapy. If you think back to those mitoxantrone and dose attacks adjuvant studies, the nice thing about an alpha is there are very few TNA-based living lifeforms that can resist exposure to alphas because of its ability to disrupt DNA.

And so you overcome treatment resistance, but yet can scavenge using a targeted molecule for circulating individual cells and render them inactive or dead with an alpha attached to it and yet minimize the collateral damage to normal organs. So I think that, right now, of course, we have therapies that, in the rising PSA context, can prolong survival and can delay progression, but can't render the patient free of disease. And as long as you can get the alpha molecule to sidle up to the cell, then you have the potential to kill that cell in ways that we don't have with systemic therapies.

Now it may be that you need combination therapy with other agents, could be that you could do it alone, we don't know. The timelines would obviously be long because it's prostate cancer, and OS is a very long end point, but if you look at failure to cure essentially as your evidence of treatment failure, I'm not sure you need to wait for OS in order to know whether you have a success or not. But if you do want progression, MFS is pretty good for those early trials and for non-metastatic CRPC, of course, we now have evidence that OS is a feasible endpoint to achieve with that so... but it probably would be an alpha.

Alicia Morgans: Well, thank you, and I just can't help but ask one more question before we finish up. PSMA varies. Expression varies over the course of disease and that's something I worry about a little bit. Can either of you comment on the changing expression over the course of disease from hormone-sensitive to continued treatment and the mCRPC setting?

Scott Tagawa: Probably both comment. I don't know that anyone knows the answer. The dogma has been that worsening disease, so higher grade, higher stage, more AR-targeted therapies, there's more PSMA until [inaudible] and multiple others neither AR nor PSMA. So that's been the kind of dogma and it's somewhat true I think on a global basis. You look at the PSMA scans almost everyone will have things that light up that may not represent- actually, no, from the Hofman slides, not all tumors will light up. So one may light up and one may not and within a tumor- look at the DeBono data- it's not all there, so we don't really know. I think it's fairly clear in cell lines that PSMA expression goes up as they're treated with hormone therapy until they die. But it's not entirely clear to me how that works in every single human.

Michael Morris: I'm glad you brought that question up because I hear many talks and read many papers that begin with PSMA as an ideal target in prostate cancer. And it really is far from ideal because its expression in some contexts is pegged to AR-signaling. Which when you look in prostate cancer that may work to your advantage or it may work to your disadvantage, depending on which way things go. There's lesion-to-lesion heterogeneity of expression which means that you are dealing with, again, differences in the delivery of radiation to each of those. It varies by disease site, by Gleason score, by neuroendocrine expression, by where the timing is in relationship to hormonal therapy.

So I think that the idea of radiopharmaceuticals is in need of a better target. PSMA is very good. It's better than anything that we have right now. But your question points out that it really isn't ideal in the sense that it's pretty variable by histologic type, by histologic grade, by whether the AR is signaling or not, by whether it's castration sensitive or resistant, or, as we seem to be learning now, by whether the patient has a DNA repair defect. So all of these are at work while you're trying to test the therapy and it would be great if we had a much more consistently expressed target than PSMA. So we have a lot of work to do in the lab to identifying a new target.

Alicia Morgans: Well, I so appreciate this discussion and I think we've learned a lot and really the enthusiasm for what we're learning about lutetium and how that may be moving into clinical spaces and the enthusiasm to discover more, whether it's more information about lutetium and PSMA targeting, or whether it's even potentially moving on to future targets, is really, really high. So thank you and I appreciate your time.

Michael Morris: Thank you and thank you, Phil, and thank you, Neal, and thank you, Scott.

Scott Tagawa: Thank you all.