Read the Full Video Transcript

Robert Flavell: All right. Nice to see you all again. I'm back. I'd like to again thank the organizers for the opportunity to present, particularly Tom, who texted me on Wednesday night to ask me to do this talk as a pinch hitter. In any case, the topic of my lecture is new targets in prostate cancer. This is an enormous topic, so I have admittedly a somewhat biased approach towards it. I'm also going to look at this with a lens focused on antibody-based radiopharmaceuticals, rather than the small molecule types of agents we've been focusing on during this meeting. Disclosures, the same as I had yesterday.

I'm going to start this lecture from a slightly different framework, which is looking at the class of theranostic radiopharmaceuticals as a whole, rather than looking at them by target. Theranostic radiopharmaceuticals can occupy a variety of different sizes and shapes. The most classic examples of this would be elements, as here in the top left, radium-223, which we're going to be talking about in a little bit, being the one that's most used in prostate cancer. Small molecule agents, such as the Glu-urea, PSMA ligands, we've been spending the whole conference talking about, peptides. These three here on the top certainly are occupying most of the space these days in clinical radioligand therapy research. But these other compounds here on the bottom line should not be totally neglected. Actually, antibody-based radiopharmaceuticals have picked up a little bit of steam recently, and I'm going to give you some examples of how that is happening.

I'm going to talk a little bit about the case for and against antibody-based theranostic radiopharmaceuticals. Well, advantages here are that there's a large variety of clinical-grade targeting vectors based on the great therapeutic success of unmodified ADCs, etc. There's reliable and modular conjugation chemistries, a wealth of preclinical and clinical data available for comparison, and high affinity and specificity. This, of course, comes at a cost, and there are some disadvantages here, particularly that full-length IgGs have a long circulating half-life. This will increase the radiation dose for the patient, for diagnostic agents, and potentially also the toxicity for radioimmunotherapy. This often requires multi-day imaging protocols. The patient gets injected on one day, comes back some days down the road. There are additional issues around storage and the cost of production compared to small molecules, as well as a potential for infusion reactions because of the antibody.

For those of you who are not as familiar with working with this class of radiopharmaceuticals, here's a nice example shared by Joe Osborne on the use of the anti-PSMA J591 antibody. This is what it looks like in imaging over a multi-day study following administration. On the early day, you see a lot of activity retained in the blood pool, and as time goes on, it gradually clears out of the circulation with some clearance through the hepatobiliary pathway. And then as time goes on, you can see gradual accumulation in metastatic bone lesions in this particular patient. Just to give you a sense compared with our traditional small molecule PSMA PET agents where we're scanning about an hour after administration.

Now, these can be converted into therapeutic radiopharmaceuticals as well. Here's some recently published data on taking the same antibody, J591, and labeling it with actinium-225. Interestingly, this compound seems to have a fairly benign side effect profile, and at least in this phase one study, they did see some PSA responses. Certainly, some enthusiasm about this approach in general of using actinium-225 or other alpha or beta-labeled antibodies.

Remember how yesterday I told you that PSMA imaging and therapy is useful across all prostate cancer patients? Well, it turns out I lied, and there's a subset of patients where it does not work quite as well. As nicely outlined by the other presenters so far today, there's a subset of patients who are not candidates for PSMA radioligand therapy. For example, in the therapy randomized phase two trial, 200 out of 291 patients were ineligible for PSMA-617 therapy. This, of course, is a very stringent inclusion criterion, but it gets the point across that not everyone; it's not a one-size-fits-all approach.

Here's an example of something that might also be seen in the clinic. This is a patient who was referred for actually actinium radioligand therapy. You can see a typical example, like some of the tumor board cases or what you might see in your own clinic, with a pretty heavy burden of nodal metastatic disease. This patient was treated with actinium PSMA, and you can see while there's clearly some response in nodal disease, this patient had an unfortunate outcome in which there was development of PSMA-negative liver metastasis. Something that we'll see in a subset of our radioligand therapy patients. We'll see resistance in some cases post-therapy with target downregulation. This reveals that there is still a need for alternative therapeutic targets in prostate cancer. Certainly, while PSMA is a wonderful target and justifiably so, there's certainly room for looking at other agents as well.

This is an enormous field. This is really just a partial list of other targets that are currently being evaluated with antibody-based radiopharmaceuticals. I'm going to focus on two in more detail in the subsequent talks or subsequent slides. There are just a couple of ones to highlight briefly here. HK2 is a nice one that's been evaluated quite a bit now with an actinium-based agent in clinical trials. And because we're holding the conference here at UCSF, I wanted to give a shout-out to the work of Mike Evans, who's been working on developing CDCP1 as a therapeutic target in prostate cancer and other malignancies as well.

Neuroendocrine prostate cancer and these advanced de-differentiated forms of prostate cancer were nicely introduced by Dr. Rettig earlier today. Essentially, what can happen in a subset of patients with advanced prostate cancer, following sustained androgen receptor inhibition, is that a subset will undergo this treatment-induced transdifferentiation to another phenotype, for example, neuroendocrine prostate cancer. These neuroendocrine prostate cancers are an aggressive variant that are typically resistant to hormonal therapy and have poor outcomes in overall survival. These are often, as Dr. Rettig said, not amenable to imaging or treatment with PSMA.

Interestingly, there are some other antigens that may, in fact, be upregulated in this patient cohort. This is some nice work here from Himisha Beltran's group. She was looking here at an antigen called delta-like ligand, DLL3. This is an antigen that turns out to be highly expressed in this cohort of neuroendocrine prostate cancer patients. In this study, it seemed that these patients have particularly poor outcomes.

Some nice slides here, courtesy of the group of Jason Lewis at Memorial Sloan Kettering. They've been looking at imaging DLL3 using this zirconium-89 SC16 ligand. You can see here clearly, in the neuroendocrine prostate cancer mice, there's quite a bit of uptake in these subcutaneous xenograft tumors. Whereas in androgen receptor-independent adenocarcinoma, there's really not much uptake. Interestingly, this antibody particularly seems to be targeting DLL3-positive tumors.

They've also converted this to a radiopharmaceutical therapy. Here, this is a lutetium-177-based agent. Interestingly, they saw really durable responses using even fairly low doses, indicating that this has some promise, in fact, as a therapeutic agent as well. They've also translated this imaging agent, the zirconium one that I showed, into the clinic, and this is an example of a patient who had metastatic prostate adenocarcinoma with biopsy-proven small cell nodal metastasis. Interestingly, you can see that there are clearly some positive lesions on the DLL3 PET, but also quite a bit more disease on the FDG PET-CT, indicating some target heterogeneity in this patient. These data were just, I looked the other day, this had just been posted on MedRxiv if you're interested.

With that, I'm going to switch gears and talk a little bit about a target that we've been looking at here at UCSF, called CD46. This is a protein that's also highly expressed in prostate cancer, and it seems that it is particularly highly expressed in dedifferentiated tumors and those with lineage plasticity, such as neuroendocrine prostate cancer, as you can see here compared against PSMA in a neuroendocrine prostate cancer cohort. There's some high expression of this antigen in PSMA-negative tumors, and there was an antibody named YS5, which was developed by phage and yeast antibody display by our collaborator, Dr. Bin Liu, here at UCSF, which targets a cancer-specific isoform or epitope on this protein.

This has been converted into an antibody-drug conjugate called FOR46. This is the results from the phase one trial. This data is courtesy of Rahul Aggarwal, who gave the prior talk. In this phase one dose-escalation study, they did see some PSA50 responses, with a mean duration of greater than 16 weeks. We rationalized that this same antibody could also be converted for theranostic purposes for both imaging and radiopharmaceutical therapy. We labeled it with zirconium-89 using standard DFO conjugation chemistry and performed some preclinical ImmunoPET imaging studies.

This here is in a panel of patient-derived xenograft models from the Living Tumor Laboratory, including LTL331, which is adenocarcinoma, LTL331R, which is a treatment-emergent neuroendocrine prostate cancer, and LTL545, which is primary neuroendocrine. In this case, we saw good uptake of the antibody in all these models. We compared it against gallium PSMA-11 PET imaging, which showed good uptake in both the LTL331 and 545 but not in the LTL331 treatment-emergent neuroendocrine prostate cancer, suggesting feasibility of the CD46-directed imaging and therapy in human treatment-emergent neuroendocrine prostate cancer, as well as potentially in PSMA-negative disease.

We've translated this agent into the clinic for PET imaging. We've scanned a few patients already. Here, this was the first patient that we scanned. We were able to see some uptake in some small lymph nodes in the retroperitoneum supraclavicular region and some bone metastases here. It's too early really to analyze this data thoroughly. We've only scanned six patients so far, but maybe at next year's meeting, I'll be able to provide an update. We've also converted this into a radiopharmaceutical therapy agent by labeling the same antibody with actinium-225. Initially, we used standard DOTA conjugation chemistry. We did biodistribution studies. As expected, we see good uptake in the tumor. And as expected also, it takes a long time to clear out of the circulation, consistent with the fact that it's a full-length IgG. We tested it in some preclinical models, including 22Rv1, which is a PSMA fairly low adenocarcinoma model. We saw pretty good responses with improvements in survival compared against control arms, and then also in this LTL545 neuroendocrine prostate cancer model that I showed you earlier. So, clear improvement in overall survival in these treatment arms.

In more recent results here, we've been tinkering around with the chelation and the linker chemistry. The initial and much of the earlier work in actinium conjugation typically used the DOTA chelator. It turns out to be a little... Not the most reactive chelator towards actinium-225, often requiring higher modification of the antibody and/or elevated temperatures. We adopted some nice recent work from Nikki Teal and Justin Wilson's group at Cornell using this Macropa chelator, which is a more effective chelator for actinium-225. We modified the chelator a bit with some new linker and conjugation chemistry and put it on the same antibody.

And then, we compared it against our DOTA as well as a few different variants of the same antibody. We found, interestingly, that one of the agents seemed to come out on top, one that had this short PEG4 linker here. We compared the DOTA-based agent against this PEG4 linker/Macropa version, and there was clearly an improvement in overall survival as well as tumor responses. I think it highlights that there's more to the development of these types of theranostic agents beyond just the antibody and the isotope. There's actually quite a bit of devil in the details in terms of how you conjugate and quite a bit of room for improvement there in our field.

With that, that'll bring me to the conclusions of the lecture. ImmunoPET and radioimmunotherapy offer potential advantages for cancer imaging and treatment, owing to a large selection of targeting vectors, often having ready access to GMP material, and high affinity and specificity. There are a number of cancer-associated antigens in addition to PSMA that are being explored in preclinical and early clinical trials in prostate cancer with a lot of promising results. Today, I highlighted work with DLL3 from the Memorial Sloan Kettering group, and CD46, from here at UCSF.

I'd like to especially acknowledge the people in my laboratory who did the work that I presented. Anil Bidkar, who led the actinium DOTA work; Sinan Wang, who led the zirconium antibody work; and then Naidu Bobba, who led the new Macropa chelator work. I'd like to thank Jason Lewis and Joe Osborne for the slides. Thank you for your attention.