Read the Full Video Transcript

Martin Pomper: Thanks again to the organizers for inviting me to talk a little bit about some new-ish stuff. Some disclosures.

So as I think we can all see, the PSMA space is getting crowded. Nevertheless, the fact that we're not seeing a huge number of cures, that I can tell, means we can still innovate—in terms of radionuclides and isotopes, as we've just seen, targeting platforms, precision dosimetry, and rational combinations, in particular adding some sort of immunological component, so that we can get to the state where no clone is left behind.

I first want to just touch upon the time-honored 617 scaffold. So we've been hearing about actinium-617, which is more or less coming online. It seems to be doing quite well, and it seems to be at least comparable or better than lutetium. In fact, if you look at the hematologic toxicity, it seems like it might be better, although we'll have to get a little bit more experience with it.

It does come with that xerostomia, but considering some of the trials that we've seen—where patients have already had lutetium and pretty much everything else thrown at them—the actinium patients do quite well. So that's coming online soon.

Now I just want to compare it a little bit to cabazitaxel, to remind us—remind myself—that we really need to try and talk to our medical oncologists and the other providers who are managing these patients. Not all of them are like Michael Morris and Emmanuel Antonarakis. And I've had a few that have told me, “Oh, I will never use those molecular therapeutics. They just don’t work.” But they do work.

And you can see here that at least with the PSA50, 617 stacks up pretty well against cabazitaxel. And it also doesn’t have nearly as much neutropenia, diarrhea, back pain, alopecia—which are the reasons that patients really sort of clamor for the molecular radiotherapy. But it does have that xerostomia, which has to be dealt with.

So a few years back, we synthesized this compound and analogs of it, and it seemed to work about as well in this preclinical model as the 617. The actinium version worked better than the lutetium version. And very recently, the gallium version has been given to patients, and you can see that there’s very little salivary gland uptake, although there’s still pretty good uptake in that bone met in the spine. This is currently undergoing a trial overseas, and I look forward to seeing how that turns out.

But before we were worried about salivary glands, we were worried about kidney. And we’ve been hearing a lot about kidney. So with Michael Zalutsky, we synthesized a compound that had a renal brush border cleavable linker. So there is PSMA in the proximal renal tubules, so we do get specific binding there, although it’s off-target. And maybe we could cleave off the activity so that it doesn’t stick around in the kidney and cause problems.

So this compound did show lower uptake in kidneys, a higher tumor-to-kidney ratio than a close analog, but it had lower tumor uptake. So we still need more medicinal chemistry around that if we want to pursue it.

OK. So I just want to talk for a minute about an astatine-labeled compound that we’re developing. This is going to be covered in more detail later by others. But there are several common alpha emitters. Astatine-211 isn’t quite there yet, but I like astatine because it has one alpha particle that comes off per decay. So after the initial decay—where there could be a recoil effect—you lose track of the alphas that come off of actinium, which could cause off-target problems.

Astatine just has one decay. But equally important is that it’s a halogen and behaves like iodine. You can use the same chemistry for it. You can incorporate it into small molecules without changing the pharmacokinetics of the molecule like you would with a chelator. And it’s got a fairly short physical half-life, which has pluses and minuses—but I’ll try to focus on the pluses. It also provides high linear energy transfer (LET) with a simpler decay scheme.

And I want to mention that the polonium decay does produce an X-ray that can be detected with a gamma camera. So it’s a visible isotope, directly visible. Actinium is not visible, and lead does have a high-energy daughter that has to be contended with. Also, the availability is up and coming. I think the capacity is increasing in the U.S., Europe, and Asia.

So here’s the compound that we made. This we considered a third-generation compound. It was synthesized by Yutian Feng and Michael Zalutsky’s group, and it compared favorably to a second-generation compound in a double-label experiment shown below. And if you look on the right, you can see that we get tumor uptake out to about 21 hours, which is three physical half-lives for astatine-211, which matches the PK of the small molecule.

So they match. We don’t really have to have it stick around longer than that. So this might enable us to compress the dose interval that we give—if we give multiple doses of the compound—and maybe give higher upfront doses, which is another potential advantage of astatine.

So this has been given to a few patients. And you can see uptake in that iliac bone. And you don’t really see much salivary gland uptake, because we used that similar scaffold to the one that doesn’t have a lot of salivary gland uptake. On the right is another recently published astatine compound.

Also, I just want to mention another thing about 617. I like studies like this. The same people who developed 617 are now trying to modify it a little bit. So a little bit of medicinal chemistry here where they want to make it more hydrophilic so that it doesn’t bind to serum albumin and—potentially on that basis—binds less to non-specific sites.

So they switched out the naphthyl and put in a styryl group instead. And they also took the cyclohexyl and made it into a phenyl. So if you look at the P18, the phenyl may have caused a little bit of extra spleen uptake, if you can see the blue on the left—but the P17 looks pretty good. And by switching out the naphthyl for styryl, and having it a little bit more hydrophilic, that will enable you to use a more lipophilic chelator. So you could potentially expand your array of isotopes, which I thought was a very nice idea.

But if you look at the overall activity of these compounds, it’s really hard to beat 617. So despite that med chem, you have things very similar to that. But they wanted to do that to decrease nonspecific serum binding.

This is the opposite approach, where they’re trying to get binding to albumin. And I just want to acknowledge this approach because it’s a common approach now with next-generation compounds. The idea being that you hitch a ride onto albumin, get long circulation, more uptake in the tumor.

What this group specifically did was they took the AMBA, which is a more lipophilic moiety, and put it—if they put it closer to the albumin-binding portion of the molecule—they’ll get more albumin binding, more tumor accumulation, better preclinical results.

However, that long circulation time also provides more exposure to normal tissues, particularly marrow, which I would be concerned about. I think the jury is still out on the utility of this approach. And to my mind, I feel like if you’re going to do this, you might as well use one of Scott’s antibodies, which also has a long circulation time—and they seem to be getting good results with those.

There’s also a heterobivalent approach. One of the irritating things about PSMA: it’s not expressed as much in neuroendocrine prostate cancer. FAP is expressed in neuroendocrine prostate cancer. So if we could target both at once, we might be able to hit both types of cancer. So we’ve tried that approach. We’ve made lots of different compounds for many different modalities.

But really I just want to call out for a minute—even though this paper is 10 years old—I’ve always really liked it. This is David Spiegel’s group. In a particularly brilliant stroke, he took two PSMA-targeting moieties, envisioned those as being Fab fragments of an antibody, and then attached those to an Fc effector. So it was a totally synthetic antibody that he used to potentially bring immune cells into the prostate tumor, which is a notoriously immunologically cold tumor. So there’s a lot of creative things that are still out there to be done—I guess that’s the point behind that.

And last, I just want to highlight Kevin Leung, who I believe is somewhere in the audience. Kevin was a BME grad student at Johns Hopkins. Now he’s an assistant professor there. And he has developed an AI-based, fully automated way of detecting, localizing, segmenting lesions, extracting various features from them, and then hopefully eventually combining that with the protected health information that’s increasingly available from the electronic health record, so that we can make extremely specific, personalized recommendations to patients.

And the idea behind that is that we want to combine the power of artificial intelligence with the sensitivity and specificity of molecular imaging, so that we can get as close to human biochemistry and molecular profiling non-invasively as possible right now.

So that’s the idea. And that is all.