Read the Full Video Transcript

Oliver Sartor: Hi, I'm Dr. Oliver Sartor with UroToday. And it's a pleasure for me to welcome Shahneen Sandhu, I should say Professor Shahneen Sandhu, at the Peter Mac down in Melbourne, Australia. And just for a brief intro for those who may not know her. Shahneen has done some amazing work with combination therapy with lutetium-177 treated-patients. And guess what? She's at it again. So Shahneen, we're going to talk about the EVOLUTION study that you presented at ASCO, and this is an interesting study looking at a couple of checkpoint inhibitors in combination with PSMA-617 lutetium. So tell us about your study and why it's important.

Shahneen Sandhu: Thank you, Oliver, for the introduction and also for the invitation to share some of this data with the broader community. I think I might share some slides which might make the details about the study a little bit more clear. So this is the schema of the EVOLUTION trial, and this was a investigator-initiated phase 2 clinical trial that was run across 10 sites in Australia, leveraging the infrastructure that we've set up in this country to do these sorts of investigator-initiated trials. The EVOLUTION trial was sponsored by ANZUP, and it was funded by Prostate Cancer Foundation Australia in conjunction with our pharma partners, Novartis and BMS, that provided additional funding support in addition to drug. The premise for this body of work was, we completely understand that radioligand therapy, particularly targeting PSMA, is highly exciting. We've seen with lutetium-PSMA-617 that it improves both progression-free survival and overall survival and quality of life, and is currently in the clinic being used for patients with metastatic castration-resistant prostate cancer.

And it is likely that these therapies move forward in the context of the treatment paradigm for prostate cancer. One of the challenges, of course, with prostate cancer is, unlike other cancers where immune checkpoint inhibitors have been a great game changer with many patients living long periods of time and the disease being under excellent control, immunotherapy in prostate cancer has really been very modest in terms of clinical benefit. And the premise for this body of work was to try and see if we could leverage the immunogenic potential of radiotherapy, particularly radioligand therapy that is directed towards multiple tumor sites at the one time. Therefore, perhaps affording the release of tumor antigens, modulating the microenvironment across multiple sites at the one time, and creating the immunogenic response, and then leveraging that immunogenic cell killing that was occurring across sites with a backbone of immune checkpoint inhibitors to see whether we could actually improve both the depth and the duration of response. So that was the backbone of this body of work.

We set out to do this with the key eligibility of men with metastatic castration-resistant prostate cancer who had been exposed to at least one androgen receptor pathway inhibitor. We wanted patients who had a decent ECOG performance status with zero to one. And we had some imaging criteria with evidence of PSMA expression there, PSMA SUVmax of greater than 15 at one site of disease and greater than 10 at other sites of disease, and patients with stratified based on prior docetaxel given there was some data to suggest that in fact the patients who had not had prior docetaxel might do better than the ones that have had docetaxel, perhaps by virtue of more advanced disease and more heterogeneous disease.

Patients would then randomize two to one to the experimental group, which was made up of lutetium-PSMA-617, 7.5 gigabecquerels every six weeks for maximum of six doses in conjunction with induction ipilimumab and nivolumab. Ipilimumab 3 mg/kg every six weeks for four doses, plus nivolumab 1 mg/kg every three weeks for eight doses. So that was the induction component, followed then by the standard maintenance strategy with a fixed dose of nivolumab. The control arm was standard dosing of lutetium-PSMA-617, 7.5 gigabecquerels every six weeks. Primary endpoint of the study was 12 months PSA-PFS. We wanted to have a landmark endpoint, given this is immunotherapy that we were evaluating. And we also at the time of designing the study, the study was being reported out. So we had a sense of what a 12 month PSA-PFS looked like for the single agent, and so we were able to benchmark the statistics for this study against that at the time of the design of the study. The secondary efficacy endpoints, as you can see highlighted here, PSA response rate, adverse events, radiological progression-free survival, and overall survival, as well as quality of life.

Oliver Sartor: Shahneen, let me ask a couple of questions if you don't mind, just for clarification. So number one, commonly in Australia, you'll use an FDG PET and have the mismatch criteria that was utilized, say in therapy. Was there an FDG PET selection criteria here in EVOLUTION?

Shahneen Sandhu: We followed the same approach, Oliver. Patients did have a FDG PET scan, and we did exclude patients that had discordant disease. So that was a component of the inclusion criteria.

Oliver Sartor: Got it. And that would excluded what may be 20% of patients or thereabouts. How many patients?

Shahneen Sandhu: I think it was about 15 to 20%. I don't know the exact number off the top of my head this morning.

Oliver Sartor: That's okay. Close enough. And then another question on the ipi-nivo. It's a little bit interesting about whether or not you use three and one, as in 3 ipi and 1 nivo, or whether or not you go the other way around and use 1 ipi and 3 nivo. And I was just a little curious about the selection of 3 ipi, 1 nivo as opposed to the other way around. Just want to hear your rationale briefly.

Shahneen Sandhu: Great question, Oliver. The basis of this at the time we were designing the study, the CheckMate 650 study that was evaluating two dosing combinations, as you say, ipi 1 or ipi 3. There is also a bit of data in other tumor types that a six weekly schedule with ipi 3 is reasonable and has the same sort of toxicity profile as ipi 1, but with a three weekly schedule. My thinking is that certainly there is data both in melanoma, but also in prostate and other tumor types, where actually dose of ipi matters. And I was thinking, here we are dealing with a highly immune-suppressed tumor, and so therefore it made some rationale to me to try to go with the higher dose of ipi. Additionally, the scheduling of six weekly also seemed to fit in very nicely for pragmatic reasons with the lutetium-PSMA-617 dosing schedule.

But then also a little bit of biology into that was, we know that radiation can actually join MDSCs (Myeloid-derived suppressor cells) and also T-regs are implicated as well. So my thought process was perhaps that scheduling where you've got the radiation load impact coming in, but you're trying to exploit the tumor microenvironment, or that hot tumor microenvironment, but that may also be a little immunosuppressed with the addition of ipi in that schedule. Does that make sense? There was no concrete signs, but it was a combination of a lot of things that made sense, as well as the fact that at the time, BMS were exploring this scheduling versus the ipi 1 scheduling as well. And I had some internal data from them that, in fact, the tox profile was fairly consistent across the two groups.

Oliver Sartor: Good. I'm sorry, one more question because there's also a little bit of discussion about the exact timing of the IO versus the isotope. When you are laying this out, is this like a same-day administration of the ipi-nivo and lutetium, or have you staggered it slightly one way or another? Just help me understand that exact timing because this is a discussion point among those who are doing combination therapy.

Shahneen Sandhu: So there is some preclinical data that suggests having immune checkpoint inhibitors on board ahead of radiation is helpful. There is some retrospective melanoma data that suggests that the response rates in the subset of patients that have radiation in the resistance setting, in conjunction with the backbone of immune checkpoint inhibitors, do better when the immune checkpoint inhibitors are already ongoing ahead of the radiation versus the radiation first followed by immunotherapy. The basis of this was partly a little bit of biology, but to be honest, not robust biology to underpin this decision. And then also pragmatic issues. I did not want any issues from the nurses and the day units where immunotherapy was being infused, if there was any concern that patients were exposing nurses to radiation or any of those sorts of concerns. And so, therefore, to deliver the immunotherapy ahead of the radioligand therapy made sense to me. And so there was a pragmatic element that was in the consideration, and then also a little bit of retrospective data and some preclinical data.

Oliver Sartor: Okay. So when you say ahead, is it one hour ahead, one day ahead? How much ahead?

Shahneen Sandhu: Yeah, so we had a window, I think it was four or five days. So you could certainly do it on the same day, but it was definitely... It had a window, and often patients were having it three to four days to five days ahead of the radioligand therapy.

Oliver Sartor: Very important because again, this has been a discussion point among those who are looking at combination with IOs. Thank you for the clarification. Yes, tell us a little bit about what you found, and it's quite interesting, I think.

Shahneen Sandhu: Thanks, Oliver. I suppose one thing I just want to highlight was we had planned to recruit a hundred patients, but we in fact closed recruitment at 93 patients. What happened was we saw one case of grade 5 myocarditis. The circumstance, there was a delayed presentation to the hospital, and this gentleman unfortunately died. As a consequence, we had a dear investigator letter. And on the back of that, there was a bit more vigilance around myocarditis events. We know that we do see myositis and things like that. And so there was an increased awareness and looking for the possibility of cardiac myocarditis, essentially. So we halted the study when we found four cases of possible myocarditis. One, as I said, was a grade 5 event and the rest were grade 2, 3 events. And I can certainly go through those in a bit more detail if needed.

Essentially, the trial management committee felt, "Look, we had recruited 93 patients, we've seen one death and some other events, and maybe it would be prudent to halt at this point in time." So the decision was made in conjunction of where we were at in the study, plus some of the side effects that we were seeing. And the trial management committee at that point felt, given the toxicities recurring in the induction phase, anyone on induction, particularly that ipi-nivo induction, should stop that. However, if anyone had passed that period and were now on the maintenance component of their therapy, then that would be permitted. They would be permitted to continue that at the discretion of both the physician and the patient. And so we went back and discussed it with each individual patient. And many of the patients were getting the benefit of the treatment and opted to stay on, and some patients opted to come off. And so when we report the data, it's a modified ITT analysis with 57 patients assigned to the lutetium-PSMA and ipi and nivo, and then 30 assigned to the lutetium-PSMA alone.

In terms of baseline characteristics, these were pretty standard for men with metastatic castration-resistant prostate cancer. All of them had had prior exposure to an ARPI, as was the inclusion criteria, and 80% of patients had had prior docetaxel. As I said, the primary endpoint was PSA-PFS at 12 months, and we saw a benefit in this primary endpoints. The PSA-PFS at 12 months was 33% for the lutetium-PSMA-617 and immune checkpoint inhibitor group, and then it was 17% for the single-agent lutetium-PSMA-617. The median PSA-PFS was approximately seven months, as you can see there, and the hazard ratio for PSA-PFS was 0.71. When we looked then at radiological progression-free survival, again looking at a landmark time point of 12 months, 27 out of 57 patients had not progressed.

So 47% of patients had not progressed on the combination arm and 23% of patients had not progressed on the single-agent lutetium arm. And the hazard ratio for rPFS was 0.53, as highlighted there. The median rPFS was 12 months for the combination arm and 8.5 months for the lutetium arm. And then I suppose in terms of responses, and I've reflected on this, I think heavy lifting is clearly being done by the radioligand therapy, but you do see a deepening of the responses here. So 50% response rates were 75% in the combination arm versus 67% in the single agent arm. PSA 90 46% versus 43%. But then, if you look at really the subset that are getting really deep responses, and many of these patients actually have not been treated for a number of years now, that's 19% versus 7%. So more patients essentially having those sorts of deep and meaningful responses.

And again, that translates into what we were seeing both from an objective response rate in the subset that had RECIST measurable disease. So 71% for the combination arm and 50% for the single-agent lutetium arm. As you can see, a number of these patients were able to suspend treatment because of that deep responses. We had a pre-specified criteria for suspension of treatment, which was a very low PSA of less than 1 and no evidence of PSMA avidity on the SPECT scans that we were undertaking post-treatment, essentially and we found that we were able to suspend treatment in 16% of patients on the combination arm and no patients on the single agent arm.

Oliver Sartor: Shahneen, let me ask you just briefly. Now, you've also looked at another IO, pembrolizumab, which has many analogies, of course, to the nivolumab. Do you think there are distinctions between the ipi-nivo approach as compared to the Pembro approach? And if so, what distinctions might be apparent to you?

Shahneen Sandhu: Also, Oliver, this response rate and these PFS benefits and the similarities is not lost on me as well, and you have to think about it in the context of the additional toxicity as well. The fact of the matter is that we would need to do a deeper dive because these are small... Even though this is almost a hundred-patient study, it's still small, and there can be a lot of nuance in that. We know that if somebody has got smaller volume disease, if they've got predominantly nodal disease, they have prognostic features that bode well on its own and might in fact be impacting those sorts of deep responses and durable responses that we might see.

I think it's hard to be certain in terms of the definitive contribution of ipilimumab, which is the key question being asked here, and it's certainly one that I'm interrogating in the context of paired biopsies and biospecimens. One of the things I would say is, it is interesting to see that the suspension of treatment rates of 16%, the PSA levels that were undetectable in 19%, which I feel is higher than what we saw with the single agent. But my view about these sorts of therapies, it's clearly a subset of patients that are getting the most value from this sort of approach. It behooves us to really try to understand that who benefits from this sort of therapy.

Oliver Sartor: I know at ASCO in the question and answer, you commented about the importance of the biomarkers and how you had reserved significant commentary about patient selection to do that deep biomarker dive. I thought that was a very appropriate answer given the fact that some patients appear to have substantial benefit, but as of yet, we're unsure of who those patients might be. Obviously, it's not going to be the MSI-high deficient-MMR type patients because you don't have enough of those. That's only about a 3 to 4% incidence anyway. You're not going to find that type of predictive biomarker yielding this type of result. Anyway, it will be interesting. In fact, I get to ask you because you're on UroToday. Any timeline for that biomarker analysis, because I think it could be quite informative?

Shahneen Sandhu: The work is ongoing. We are starting with simple things like RNA signatures. And I think there is a little bit of data around interferon signaling and how important that is both in terms of durable responses to radiation as well as across tumor types, but then also in the context of immune checkpoint inhibitors. Those are simple things that we can actually achieve quite quickly. So we'd be looking at that and then layer that on with some spatial type analysis. But my view is, for a biomarker to be useful in the clinic, it really has to be a pragmatic and a simply applied biomarker. So, yes, we can do lots of deep dives and look at things from a spatial perspective and have all sorts of multiomics, but ultimately, how is it going to enter into the clinic? So we've got to simplify it. So that's going to be the approach with this study. So, in terms of timeframe, I would expect 12 months. So the work started, and we've got two people working on it. So my expectation is approximately 12 months.

Oliver Sartor: Thank you. All right. I think you had one more slide you wanted to share on the toxicity issues?

Shahneen Sandhu: Yeah. It's important to think about immunotherapy or radioligand therapy, any therapy for patients in the context of clinical benefit and risk. And so this volcano plot, of course, highlights the impact of immunotherapy. Overall, our impression from the side effect profile was one could distinguish between side effects that were radioligand-related versus side effects that were immune-mediated. Although we noticed four cases of myocarditis, if one looks at the rates of other immune-related toxicities, they were fairly consistent with what you would expect from this dosing schedule of ipilimumab and nivolumab. I suppose that's my takeaway from the adverse events. But the bottom line is, it's an effective therapy in a subset of patients, and we just need to understand who that subset is.

Oliver Sartor: As you additionally follow these patients with clinical endpoints, it could be quite provocative. And I would say the same, by the way, for your study with the pembrolizumab. A lot of the information is actually in the duration of the follow-up, where people are questioning will these low PSAs translate into long PFS and survivorship. As of yet, you haven't reported yet on the pembrolizumab study, but hopefully, we'll see that at an upcoming conference. And here, of course, you don't have the opportunity to present long-term data because you don't have long-term data yet. But anyway, just briefly commenting about your view of durability at this point, and why not branch it out to pembrolizumab protocol as well as the ipi-nivo protocol.

Shahneen Sandhu: Yeah. That is the plan, Oliver. We will have two years of follow-up on the EVOLUTION trial in August of this year. So around the corner. And the intent would be to present that data, hopefully at GU ASCO next year. I think with immune checkpoint inhibitors, certainly with ipilimumab and nivolumab, and of course, I'm reflecting on other cancers as well. The landmark at two years is actually quite meaningful. And so I think that's going to be a good time point to look at the durability. Certainly, we have had patients have only one or two doses of radioligand therapy and have not needed anything more.

Oliver Sartor: Is it fair to ask about your Pembro study, which now is in a longer-term follow-up? I'm not trying to put you on the spot, but if it's fair to discuss, it'd be great to have our audience here.

Shahneen Sandhu: Absolutely. So we have had longer follow-up in those patients. I think even in the patients that have had meaningful responses, and this is where that contrast between pembrolizumab and ipilimumab and nivolumab, or really CTLA-4 and PD-1, is going to be important. Because certainly, with PD-1 alone, we did see durable responses, but I can also tell you, all the patients have now progressed with three years of follow-up. So are we achieving those sorts of meaningful responses which we want to achieve for a subset of patients with prostate cancer, where we can walk away and say, job done, go and enjoy life? Unfortunately, with single agent, with PD-1 and lutetium-PSMA, we are improving the responses and duration of responses, but not sufficiently to... Ultimately, we want cure for our patients really. So that question around contribution of CTLA-4, I think, is important.

Oliver Sartor: Shahneen, I have tremendous admiration for the desire to cure. These are really tough patients. These are VISION patients. They are post-ARPI, 80% post-docetaxel. I mean, I love your modesty, but my goodness gracious. If we're going to start curing, I think we'd have to go earlier. Anyway, amazing, amazing answer.

Shahneen Sandhu: No, look, I completely agree, Oliver. I suppose we should have a lofty goal, but the... I think when we are looking at immunotherapy and these sorts of strategies that might offer patients cure, you're completely right, the space to go is very early. But I think one of the things that becomes important is we've got so many effective therapies for men early, and so we can't do harm. And so we need to fully understand in the context of mCRPC really who's getting that sort of mileage from the therapies and weighing up the cost-benefit for individual patients before we go early. I think there's value.

For me, one of the challenges with all that we've done in trying to convert this highly immune-suppressed tumor to make it more immunogenic is I think we stop shy of... We make the cold tumor hot, we definitely do that, but we also raise the bar, in that, there's adaptation and it becomes a bit more immune suppressed perhaps and we stop shy of doing something more. And my thought process is that we may need to start thinking about how we target that tumor microenvironment a little bit better. There are agents that are coming that may be in the position to do that in conjunction. So the concept of layering things on, treading lightly, and understanding biology is key.

Oliver Sartor: Without a doubt. I feel like there's a biology that we don't understand, and perhaps your biomarkers will inform. And so over the next 12 months when those analyses are being performed, I'll be looking forward to your presentation. I guess that'll be probably in 18 months when we'll learn about what you found. Shahneen, thank you so much for being on UroToday. I always appreciate the discussions that we have together, and I always appreciate your innovation and your thoughtfulness about how you take on the patients, these tough problems, and contribute to our understanding of the disease. So thank you for what you do.

Shahneen Sandhu: Thank you, Oliver. And thank you very much for having this interview and allowing me the opportunity to present some of this data. Thank you.