PSMA PET in the Metastatic Castration-Resistant Prostate Cancer Setting - Oliver Sartor
November 16, 2021
A. Oliver Sartor, MD, Professor of Medicine and Medical Director, Tulane Cancer Center; C. E. and Bernadine Laborde Professor of Cancer Research, New Orleans, Louisiana
Pain and Symptomatic Skeletal Events Reported in the VISION Trial – Oliver Sartor
ESMO 2021: Health-Related Quality of Life, Pain and Safety Outcomes in the Phase 3 VISION Study of 177Lu-PSMA-617 in Patients with mCRPC
The Clinical Implications of The VISION Trial, PSMA-Targeted Radiotherapy in Metastatic Prostate Cancer - Michael Morris
ASCO 2021: Phase III Study of Lutetium-177-PSMA-617 in Patients with Metastatic Castration-Resistant Prostate Cancer (VISION)
Oliver Sartor: Thank you for joining PSMA-PET Academy. I'm Dr. Oliver Sartor, and it's my pleasure to present the VISION trial. I'm going to talk about some of the design considerations in addition to the data, and maybe a few thoughts that when we move forward. Now, let's start off by looking at some of the goals and caveats surrounding the VISION trial. The goal was to design a pivotal trial for global regulatory approval and ensure that we have a label that is suitable for implementation in multiple geographic settings. The goal was not to design a trial to show the maximum activity of PSMA-617. The goal was to be very practical. We wanted overall survival endpoint from the start. And by the way, rPFS got added a little bit later, and that created some uncertainty at multiple levels. But the goal was to hit this overall survival endpoint by targeting a population that had few alternatives.
We wanted to keep the PSMA selection very simple and very real world. We knew that FDG PET was not reimbursed in the USA. And if we had included FDG PET, it would've been a real problem in terms of the label and selection moving forward. We also wanted to create a precedent and a paradigm that we could move forward into earlier stages of the disease, somewhere to Cougar-301, Cougar-302 in latitude for abiraterone. Now, I'm going to give a shout out to Chris Parker, and he was very involved. And I was a co-investigator with Chris on the ALSYPMCA trial with Radium-223. And we actually used the ALSYPMCA trial as a bit of a template. First of all, we wanted an image-based biomarker used for patient selection. And then we wanted to exclude certain patients who had non-targetable lesions, and that's very important. We did that in VISION as well.
We wanted an overall survival point when rPFS was uncertain at the time of trial design, and we wanted standard of care plus or minus an experimental agent. Chemotherapy was not allowed as a standard of care, and that was true for ALSYPMCA as well. So all the patients who came on division had refused chemotherapy or be unfit for chemotherapy. And by the way, if you didn't want to participate in the trial and you thought the chemotherapy was important, then go do chemotherapy. You didn't have to participate in the trial. We wanted to adopt some risk mitigation strategies, and one of these was a loose PSMA selection criteria. We did not attempt to optimize PSMA selection. We just wanted to have a metastatic lesion more than the liver, but that was a broad inclusion criteria. And we thought that this broad market opportunity would, in the end, be very attractive and would provide the most opportunity to benefit patients.
There were unknown effects of the Lu-177 PSMA-617 on bone scan progression. That's why we wanted the OS input. And only later did this rPFS endpoint get added. We could have put in chemotherapy as a potential active comparator, but the truth is, only a small percentage of men ever received a second chemotherapy. And chemotherapy, if we had combined it with Lu-177, would've been problematic. So what we did was allow two lines of a prior taxane chemotherapy. And that way we didn't have to worry about a third line being involved in the control arm, because the truth is, that there are really only two lines of chemotherapy that are used and even two lines are not very commonly used. The truth is that most men do not receive even one line of chemotherapy, and no more than about 20 to at most 25% of men will receive two lines or more.
Now, here's The New England Journal, originally online, June 23rd, 2021. Here's the basic trial design, sort of the guts, if you will. Progressive metastatic CRPC, PSMA-positive for metastatic disease with a Ga-68 PSMA-11 PET scan, had to be uptake greater than the liver. And then some negative selection criteria I'll come to in a second. You could have had up to two taxanes, but required to have at least one taxane, were required to have at least one of novel hormone. And this was typically abiraterone or enzalutamide, but you could have had two or even more. ECOG performance status, 0-2, life expectancy of greater than six months. Patients randomized two to one and you could receive up to six cycles of the Lu-177 PSMA-617 plus standard of care or standard of care alone. And in the end, it turned out with an alternate primary endpoint of rPFS and OS along with a variety of key secondary endpoints.
There were also stratification factors, such as LDH and liver metastases, performance status, and inclusion of the novel hormones that standard of care at the time of randomization. Here, we look at the selection with PSMA PET, and it turns out of the 1,179 patients assessed for eligibility, that 1,003 actually did have the PSMA PET. And of those who had the PSMA PET, 87% met the PSMA PET criteria for involvement in the trial. And again, I mentioned you had to have a PSMA positive metastatic lesion uptake greater than the liver. There was no size criteria for the PSMA positive lesion. If you had a PSMA positive lesion that appeared to be 0.5 centimeters, that was considered eligible. Now, no PSMA negative lesions greater than one centimeter in the visceral or lytic bone lesions greater than one centimeter. So PSMA negative there would exclude you or a PSMA negative lymph node of 2.5 centimeters or greater.
So that way, we had this negative selection criteria as well as the positive selection criteria. But about 87% actually did meet the PSMA PET criteria. Now, here are the patients that were actually treated and they're going to be two different subsets. I'll explain that here in a second. There's going to be a image-based progression-free survival, or rPFS subset, with 581. And then all patients, which was 831. And here, you can look at the characteristics. About 12-13% had liver mets. Around anywhere from 8-10% had lung mets. Bone mets, of course, present in more than 90% of cases. Medium PSA is somewhere in the 70s to 90s. Alkaline phosphatase is in the 90s to 108. LDH is anywhere in the 220 to 230 range. Now, this is important because, first of all, we're going to look at previous androgen receptor targeted therapy, this would be abiraterone enzalutamide. Somewhere around 45-55% of the patients had one regimen.
But somewhere between 39-45% of the patients actually had two, and a smattering of patients, anywhere from 5-8%, actually had more than two regimens. So these patients had lots of novel hormones and they all had previous treatments with taxane. One regimen was only present at anywhere from 52-59%, two regimens, anywhere from 39-47%. If we look at Docetaxel and Cabazitaxel, almost everybody had Docetaxel. It was allowed if you had another taxane, including Cabazitaxel, as first line, which is rarely used. But take a look here, and that is anywhere from 38-43% of the patients had Cabazitaxel too. I say this to emphasize, these are very, very heavily pretreated patients, multiple hormones and multiple taxanes. Now, shortly after accrual began, dropout problems became immediate in the control arms. These were predominantly sites where nuclear med docs were leading the trial.
Naturally patients were disappointed, and they did not want to get the control arm. They came to get the PSMA Lutetium. As a consequence of some problems, we had to enforce some significant issues related to the trial and accrual sites. Number one is if there was problems with the site, meaning that the control arms were dropping out, as seen in the early on 56% of the patients were dropping out, we kept enrollment at those sites. We then went back to the other sites that were doing better and said, "Get serious, guys, because we're not going to succeed in the trial unless we keep the patients on study in the control arm and the experimental arm." We've then had to go to the FDA and say, "What do we do? We've ended up with this dramatic increase in early dropout."
And they said, "Get your act together. And what we're going to do is we're going to analyze survival in all randomized patients, 831. But we'll allow rPFS to be assessed only after these early dropout measures were taken." So in March of 2019, we put these measures into effect and that's why we're going to have two different numbers that have to be explained when we look at the data. Now, it turns out that both the endpoints were very strikingly positive. Overall survival hazard ratio of 0.62, confidence intervals didn't even come close to 1, p-value less than 0.001, strong hazard ratio for overall survival and even stronger for rPFS, hazard ratio of 0.4, confidence intervals not even close to 1, less than 0.001 on the p-value. If you look at the rPFS, you more than doubled the rPFS. Looked at survival, went from 11.3 to 15.3 months on medium. By the way, if you looked at the [inaudible 00:10:22] subset for rPFS, hazard ratio of 0.43. Looked at the rPFS subset for overall survival, hazard ratio of 0.63.
Doesn't matter which way you cut it, it's all positive. Now, next looking at resist criteria, it turns out that CRs, complete responses, were seen in 9% of the patients versus 0, and PRs 41.8 in the experimental arm versus 3.1 for the control arm. You add those two together, the 41.8 and 9.2, and basically you come out with an over 50% objective response rate. That is really, really strong. Then you look at the PSA and what you'll see is PSA confirmed decrease in the experimental arm with the Lutetium is greater than 50% in 46% of patients and greater than 50% in 7.1%. Again, a dramatic difference. If you look at the exposure, and this is important, if you look at the exposure, there's much more exposure seen when you look at the PSMA Lutetium. When we look at adverse events, certain adverse events are going to be exposure-related, and this is important to member.
So now, when we look at the adverse events, clearly things like fatigue are more common, but we have a longer duration of exposure. And that may explain part of it, but there could be more fatigue in the Lutetium arm. And I think there likely is. Dry mouth unequivocal, you get more dry mouth when you use the PSMA Lutetium. Some nausea, a little bit more common, anemia, more common, grade three, four was about 12%, 12.9% with the Lutetium arm versus 4.9%. And then a whole variety of other things that are of questionable importance. Although I think you can get some diarrhea, and I think vomiting can be rarely seen. If we look at the thrombocytopenia and leukopenia, 7.9 grade three, 2.5 for the leukopenia grade three, and that is higher than in the control arm. So some bone marrow suppression, dry mouth, anemia, and some GI side effects and nausea and fatigue can be seen with this age.
So what do we know? I believe that the Lu-177 PSMA-617 is effective and reasonably well-tolerated in very heavily pretreated patients. I think it's even likely that the trial would've been positive without patient selection using PSMA PET. We have an overall survival of 0.62 and confidence intervals that are very strong and only excluded 13% of the patients. Mary-Ellen Taplin made this point as a [inaudible 00:13:08]. I'm going to say that nuclear medicine sites that are not well partnered with oncology had difficulty managing the control group. And I believe that multidisciplinary care is optimal, and I think that's going to be true in the real world as well. I think this therapy will be rapidly adopted after regulatory approvals and likely is going to move earlier in the treatment paradigm. But of course, that involves more trials. Also want to emphasize what we do not know.
I'm not sure what the optimal patient selection criteria when using PSMA PET is. Should we put in FDG PET? I don't really know, but I do know this. I know that the selection criteria we used in VISION turns out to be positive, and I believe we'll have a regulatory approval without FDG PET. Is this the optimal dose and schedule for PSMA Lutetium? I'm not certain, but I do know that if you do it this way, you could prolong overall survival. What about important questions like PSA progression, response, and survival benefit? I really don't know. It's going to take more analyses. What about retreatment with PSMA Lutetium and progression? Can that have a positive impact? I don't really know. We need to learn more. What about treatments in terms of standard of care? Did that really add? I'm not so sure about that either, but it's okay.
That's what we did in the trial and we had turned up positive. I think we have to look forward to synergistically combining other therapies with PSMA Lutetium. But right now, I don't know what those are. I think we have to ask questions about earlier in the stage of disease. And what about metastatic hormone sensitive prostate cancer? That's a whole different question, but nevertheless exciting. Thank you for the opportunity to present here today. I hope I've covered a little bit about VISION, not only what the results were, but why we did what we did and interpreting the results, hopefully with asking some questions that need to be answered as we go forward. Thanks again for the opportunity to present here today.