Therapy Resistance and Combination Therapies "Presentation" - Johannes Czernin

February 14, 2024

At the 2024 UCSF-UCLA PSMA Conference, Johannes Czernin explores the challenges of resistance in PSMA-targeted theranostics, advocating for personalized treatments based on dosimetry and strategic combination therapies to enhance efficacy and patient survival. Dr. Czernin calls for a shift in treatment approaches, emphasizing the need for optimized radiation delivery and innovative clinical trials to overcome the limitations observed in current practices.

UCLAUCSF video logo stroke


Johannes Czernin, MD, Nuclear Medicine Physician, Chief of the Ahmanson Translational Imaging Division, David Geffen School of Medicine, UCLA, Ronald Reagan UCLA Medical Center, Santa Monica, CA

Read the Full Video Transcript

Johannes Czernin: Yeah, it's difficult for me to believe that there will be much new for you after two days of listening to the different angles that people have about PSMA-targeted theranostics. We got interested in therapy resistance quite some time ago in preclinical studies, and then of course trying to translate it, or other people are translating it. What I'm going to do is, I'm going to show a mix of preclinical studies, different mouse models of disease, but always try to parallel it with clinical trials that then translated what has been observed in murine studies. My interests. So the issue of resistance is really important because... Oops, that was not good.

Because we have three prospective clinical trials. The first one was coming from Australia. It was a non-randomized trial, lutetium-PSMA, and showed a PSA50 response rate of 57%. Then we had the UCLA study with a PSA50 of 37%. And then of course, the pivotal VISION trial with a PSA50 response of 46%. When patients are responding, they live about 10 months longer than non-responders. This was shown in these two studies here. This was the outcome of the VISION trial, and you have seen that many times before. And I thought when I saw this result, it's actually not great because it's actually at the level of 223 radium Xofigo responses. And so that kind of prompted us to really start thinking about how we can improve things. And several talks today already mentioned the term we could really do better than that.

So I want to show you some of the reasons for resistance, and some of the arguments for why resistance can occur. So one is, of course, suboptimal radiation type, and we modeled that in animals. This is a model where we inject tumor cells into the left ventricle, that creates a model that mimics human metastatic disease quite well. And then we treated the mice with either lutetium, actinium, or the tandem treatment, actinium plus lutetium, or of course, a control group. This is the actinium curve down here. And this is then measured, the tumor burden is measured with bioluminescence. This is the actinium curve here. And actinium was clearly better than lutetium. That didn't have much of an effect at all, and this was the tandem that didn't lead to a significant improvement in tumor burden control. So that's important. And I think Matthias will talk about the tandem treatment in his talk. I would predict there's not much of an effect. Suboptimal activity.

This is an actinium trial. And here what's interesting, if you do... This is a very aggressive model. If you have no treatment, these mice die very quickly. Then we did 20 kilobecquerels and got some response. Then we did 40 kilobecquerels and got a very good response. But when you did 100 kilobecquerels, the response was actually worse, and the reason was that we encountered toxicity. So that's not good. It shows you in a way that you always have to be very careful, especially with actinium. And I come back to that in one of the next slides, of course, with the side effects profile and whether it's really worthwhile doing this. This is a human study here. And that showed if you have less than 10 gray tumor dose, then you will have the vast majority of patients as non-PSA responders. So you need to get enough radiation into the tumor, which is quite obvious.

And you need to optimize the delivery of radioactivity, which is not only dependent, of course, on injected activity but also on other pharmacokinetic parameters. Much was talked about low target or target expression, the importance of target expression already. Here we modeled PSMA-negative lesions, low positivity, medium, and high positivity, high-expressing lesions. And as expected, the ones with the highest expression had the best response to a single dose of lutetium. And here we looked at targeted heterogeneity. We did mixes of non-expressing and 25% to 75%, 50/50, 50% of the cells positive for PSMA expression, 75%, and 100% positive. And again, if you have the least heterogeneity, then the response to treatment will be the best.

And this was already translated. And the question is, does the radiation type matter in clinical studies? So again, here you have the lutetium-PSMA trial, the VISION trial here with a 46% PSA50 response rate, and an overall survival of 15.3 months and a progression-free survival of 8.7 months. This came out two weeks ago or so. It's a retrospective study from seven different countries, I believe, or at least centers. It's actinium-225-PSMA radioligand therapy of metastatic castrate-resistant prostate cancer, a multicenter and so on. This is actually a population that is quite similar to the VISION trial population. 70% of these patients had chemotherapy and 30% of the patients had Pluvicto before they had actinium. So what do we get here? We get a 57% PSA50 response rate and 73% at any decline.

But what I find remarkable is that you get exactly the same overall survival and progression-free survival as in the VISION trial. So it doesn't seem to me that this is really adding a lot of power to our treatment armamentarium. This is different, and I think it was shown already today by Oliver, this study. A retrospective study by Mike Sathekge. And with an excellent response rate, 91% PSA50 response rate and progression-free survival in PSA responders was 22 months. The question is whether it's worthwhile to do this as a treatment strategy, given that the toxicity of this approach is quite substantial.

And that was already shown, and it's just a reminder target expression was substantially studied in clinical populations. The highest SUV, mean SUV, the highest portal has the best response rate. But it was pointed out that also the lower ones here that are very similar in response rates to Pluvicto in this case, have a benefit when compared to the control groups. They do slightly better. So they are not automatically disqualified just because they have a lower mean SUV. This was a VISION sub-study. And then lesion heterogeneity, and that's an old study, a retrospective study in 85 patients where patients were classified as having high and homogeneous PSMA expressing lesions, high and heterogeneous here, and low and homogeneous. And to make this long story short here, of course, the high and homogeneously expressing patients or lesions had the best response to the treatment while the other ones clearly were worse. So heterogeneity is something that needs to be considered in the anticipation of responses to Pluvicto.

So what we have here, we can, and I think we should, optimize the radiation approach and patient selection. And that is really done by optimizing the radiation type. And I think it's completely unknown whether actinium is really going to provide a benefit. We need to do randomized studies to show that. The injected activity is critically important, I think. And it was said many times before, we are underdosing our patients. The tumor dose needs to be increased. And whether that may be documented in the ENZA-P study as a combination treatment approach to increase target expression, as Dr. Emmett showed, or whether we just achieve it by increasing the injected activity is a different question. And how much of side effects we get through this approach is also a question.
And this is really important. When you look at the patients that we get, they're all extremely advanced. They have tons of disease. I really believe that all oncologists send the patients too late. That's a consistent concern that I have. With this, I want to move on to combination treatments. And of course, in the next talk, we'll talk about immune therapy and combination treatment. But I want to also touch upon this. And the second one is DNA damage response inhibitors combining radioligand therapy or radiopharmaceutical therapies with DNA damage response inhibitors such as PARP inhibitors.

You have seen this here, that prostate cancer is immunologically cold, has a low tumor mutation burden, low expression of PD-L1, sparse T-cell infiltration, and an immune suppressive microenvironment. And when you look here, this is prostate cancer with a very low mutational burden. Here on top, you have melanoma, squamous cell lung cancer, adenocarcinoma of the lung, and so on, with an order of magnitude more mutations, and therefore much higher immunogenicity. And therefore, a much better subject for immunotherapy than prostate cancer. But previous studies, all studies have shown that external beam radiation can convert prostate cancer into an immunologically hot tumor. And that could then be exploited by hitting the immune system with immune checkpoint inhibitors. The question is, can radioligand therapy do the same? And the mechanism would be that you get increased release of tumor neoantigens, PD-L1 upregulation, and increased MHC-1 expression. So everything going towards an inflammatory phenotype of these cancers.

So what we try to do again in animals is to model this to understand the effect of lutetium-PSMA on dynamic immune cell changes and optimize the timing for future radiopharmaceutical therapy and immunotherapy combination. So we did a lot of mouse studies to do this. And we gave a single dose of Pluvicto, and then we did immune phenotyping at one hour, four hours, 48 hours, 72, 144, 260, and 288 hours to really find out what the dynamic changes are. And the reason why this is important is because if you get early upregulation of immune supportive systems, cells, pathways, then of course it's much more likely that your combination therapy will be effective. And these are just very early data here.

So we take tissue out, looked at the spleen, and with flow cytometry after four hours, that's the early time point. And we see that immune supportive functions such as M1 macrophages, CD8 T-cells, neutrophils go up while immune suppressive functions in the spleen go down. And what happens in tumors already, after four hours, it reaches high expression of PD-1, PD-L1, and a high degree of DNA damage response. And just to remind you, the DNA damage response, that aims at fixing the DNA damage and moving cells, as Dr. Emmett described before, to apoptosis. And that's of course what we try to avoid when we block these pathways. That gives us another angle to attack tumor cell liability.

So we then did an immune study again. It's a randomized mouse trial, essentially. You have controls, you give an anti-PD-1 antibody in a certain schedule. In this case, we treated with actinium and then did the combination treatment with radionuclide therapy and a PD-1 antibody. And as you can see here, after a single administration of the actinium, we get in the combination therapy a response with two sustained actually cures of disease. While the other ones have no benefit, no significant benefit from the treatment. You essentially almost double the overall survival of the mice. Now, of course, at UCSF, people are quick. This is a study by Aggarwal and Hope recently published in Lancet Oncology. In patients with progressive castrate-resistant prostate cancer on ARSI, they have more than three PSMA expression lesions, progression on next-generation hormone therapy. And just as a side note, 14% of these patients had prior taxane for castrate-sensitive disease.

And this was an interesting protocol, it had two arms or two parts. The first part was to find out what the best schedule is for part two. So there was a single cycle of 7.4 gigabecquerels of lutetium-PSMA to either 28 days before, concomitant, or 21 days after the start of Pluvicto. And it turns out that the 28 days before pembro would be the best treatment regimen. But look, when you look at this data, this is a single administration. You achieve a complete response in two patients. This is an anatomic response. Partial response in 47%, stable disease in 28%, progressive disease in 21%. And the PSA response is also remarkable because that's, again, after a single administration of lutetium. And here you get a 44% PSA50 response and a 16% PSA90 response. So the outcome of this study is very promising.

Then, of course, you have the people in Australia, as always, they're always there when we get there. And this was in 37 patients, this is the PRINCE trial by Shahneen Sandhu, presented at ASCO in 2022. And that was a heavily pretreated population, 73% with docetaxel and 100% on ARSI. Six cycles of lutetium-PSMA-617, starting at 8.5 gigabecquerels and then dose reduction per cycle by 0.5 gigabecquerels. And then combined with pembro every three weeks for up to two years. This was the primary endpoint and they had a 73% PSA50 response rate. They had radiographic progression-free survival of 11.2 months and a PSA progression-free survival of 8.2 months. So both studies actually concluded that these data are extremely promising, and that these provide the impetus for further evaluation of immune checkpoint inhibition in patients with metastatic castrate-resistant prostate cancer.

The other approach is, and they just have this one slide here on the LuPARP trial. Presented at ASCO 2023, again, by Shahneen Sandhu. This was planned to be 48 patients, it was not complete at that time, with high PSMA expression. Again, a high rate of prior taxane and ARSI. It was safe, I just want to say that. But the results, PSA50 response rate was 62%, PSA90 response rate was 48%. So pretty spectacular data. And we are waiting, of course, for the results of the study that was announced, that UCSF is already conducting a phase two or phase three clinical trial.

So I'm going to summarize here. I think we can do much better in terms of radiation delivery to overcome the limited effectiveness of our treatments. Injected activity needs to be done based on dosimetry. And I think we shouldn't look so much for the off-target dosimetry, but we should look more for the tumor dosimetry and really up the dose that we get to tumors. And I think it's interesting when you look at the population that we see in daily life at the clinic, they have a very, very short life expectancy. And I would be extremely surprised if any toxicity arises within the next two or three or four years in this patient population. So I think we can easily up the activity that we inject, given that we find a way, as Oliver suggested, to talk to the FDA or the regulatory bodies and make things a little bit easier.

Patient selection. There's a clear reason why the data in Australia looked better than in the rest of the world. And I think it's really the selection criteria. And I think something that is not sufficiently done is looking at mutational status. There is clearly a radiation resistance problem in P53 mutated patients. And we really don't screen for that, and this is something that needs to be done. Rational combination therapies would be DNA damage response inhibitors. It's not only PARP inhibitors; there are ATM and ATR inhibitors. Unfortunately, they have a pretty bad safety profile, but they're quite effective, at least in models. Immunogenicity, so exploiting the upregulation of the kind of inflammatory phenotype through the radiation approach and then hitting them with immune checkpoint blockers. Interferon signaling, that's something we work on. Again, not the best drugs, but we can probably use STING agonists. But combined with radioligand therapy, one could maybe reduce the dose of these drugs. And they're very, very potent in models.

And Louise talked about the ENZA-P trial that has probably, I would think, the most impressive results of these combinations, much to my surprise. And then I think what is really important is establishing the best sequencing. And there, the animal studies can really help because there's a timeline that the immune system is activated very quickly and you have to hit the target really early after the radioligand therapy with your pharmacological approach. So I hope that we can test combinations in earlier disease stages in clinical trials. And with this, thank you very much. I want to show you something.

These are the new people that we get at UCLA. And I haven't put them on the last slide because I was unable to copy a picture into the... I thought this is nice. This is the European meeting in Vienna. And obviously, some of them are having a pretty good time. And this is the team. This is all teamwork. And I just want to point to one thing here. It is very unusual. It's as if USC and UCLA suddenly become friends. That's almost impossible, but it happened. So thank you very much for your attention.