CONDOR: Study of 18F-DCFPyL PET/CT Imaging in Patients with Suspected Recurrence of Prostate Cancer - Michael J. Morris

May 28, 2020

Michael Morris, Oliver Sartor, and Alicia Morgans discuss the CONDOR study, the second of two prospective clinical trials designed in collaboration with the FDA to demonstrate the diagnostic performance of PyL in patients with biochemically recurrent prostate cancer. Correct localization rate (CLR), or the positive predictive value of the PyL scan, was established as the primary endpoint and the CLR found in CONDOR was 85-87%, with 64% percent of patients receiving a change in intended management of care. For 20% of patients, that management changed from a non-curative systemic therapy to a curative goal of care.


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

Michael J. Morris, MD., is the Clinical Director, Genitourinary Medical Oncology Service & Prostate Cancer Section Head, Division of Solid Tumor Oncology, Memorial Sloan Kettering Cancer Center, New York, NY.    Dr. Morris is one of the leaders of the Prostate Cancer Clinical Trials Consortium (PCCTC). This initiative is designed to increase patient access to clinical trials across the country. He is also the chair of the Genitourinary Committee of the National Cancer Institute‚Äôs cooperative group Alliance for Clinical Trials in Oncology, which facilitates testing new drugs nationwide.

Alicia Morgans, MD, MPH Associate Professor of Medicine in the Division of Hematology/Oncology at the Northwestern University Feinberg School of Medicine in Chicago, Illinois.

Read the Full Video Transcript

Alicia Morgans: Hi, my name is Alicia Morgans and I'm an Associate Professor of Medicine and a GU medical oncologist at Northwestern University in Chicago, Illinois. I have here with me today Dr. Oliver Sartor, who's a Professor of Medicine and the Medical Director of the Tulane Cancer Center, as well as Michael Morris, who is a Professor of Medicine and the Prostate Cancer Section Head at Memorial Sloan Kettering in New York. Thank you both for being here with me today.

Michael Morris: Thank you, Alicia.

Oliver Sartor: Thank you, Alicia.

Alicia Morgans: Wonderful. So we wanted to talk about the CONDOR study, which is a presentation that Michael has made at ASCO. Can you tell us a little bit about what that study was, Michael?

Michael Morris: Sure. So, as many of your viewers will probably know, previous data, primarily from Europe, have really shown compellingly that PSMA-PET imaging is a superior imaging modality for prostate cancer relative to current standard imaging methodologies. And PyL or [18F]DCFPyL in full, is a PSMA-directed PET radiopharmaceutical that's being studied to collect an evidentiary database in support of regulatory approval in the U.S.

So CONDOR is the second of two prospective clinical trials designed in collaboration with the FDA to demonstrate the diagnostic performance of PyL in prostate cancer. CONDOR specifically is for the biochemically recurrent prostate cancer patient. There was a prior study called OSPREY that looked at PyL's performance in men with high-risk localized disease and in men with both locally recurrent and metastatic disease, so CONDOR filled that gap that OSPREY flanks of the biochemically relapsed prostate cancer patient after definitive local therapy.

Alicia Morgans: This is a patient population that's really in need of sensitive imaging because it's a decision point where we're trying to decide do we intensify potentially with pelvic radiation or do we decide to move in a systemic therapy direction, do we decide to just watch potentially for some patients? So really an important piece of the treatment paradigm. So what exactly did the CONDOR study entail then? How did you treat patients on this study?

Michael Morris: Well, this is one of the most difficult patient populations to establish the accuracy of an imaging modality because this patient population by definition has negative standard scans, so there's really very little to compare to. And as well by definition, because they have either negative or equivocal standard scans, it's also often frequency to establish what's a true positive on the PET scan by virtue of a biopsy because there's really not much to see. So in order to have an endpoint that sort of met the criteria of establishing what a true lesion was on the PyL scan, there needed to be a composite endpoint created. And this again was in conjunction with the agency, and so the term correct localization rate was adopted as the primary endpoint. That's abbreviated as CLR. And essentially what the CLR is the positive predictive value, that is true positives over two positives plus false positives, with the added requirement of anatomic location matching. So we needed to create an anatomic atlas in order to do this study, in which you would look at a lesion that appeared on the PyL with a lesion in the same anatomic region that appeared on some other corresponding study that was termed the composite of truth or composite truth standard to verify whether the lesion on the PyL was a true positive or a false positive.

And that composite standard of truth was hierarchical, so the first and highest standard would be histopathology findings from either a salvage lymph node dissection or if feasible, a biopsy. But if that wasn't feasible, then informative conventional imaging, such as fluciclovine or choline, or a targeted MRI or CT, or some other imaging modality that hadn't been done by the patient's own institutional standard imaging algorithm for a biochemical relapse. Or finally, as the lowest form of evidence, a confirmed PSA response in the absence of concomitant ADT if radiation therapy were delivered to the lesion and the PSA declined by 50%.

So it's a pretty tricky set of endpoints because you have nothing to really verify what's found on the PyL as a standard, and hence the composite endpoint that I just described known as the CLR.

Oliver Sartor: Yeah. Mike, let me ask a little bit of a question about the second component of that composite, and that is positive by conventional imaging. And I understand that a number of these were fluciclovines, cholines, in addition to the targeted CTs, MRIs. But that one's a little bit tricky because it means actually you could have detected it with an alternative imaging modality that's currently available. What percentage of the patients actually fell into each of those three categories? I mean the pathology's tough to obtain, the conventional imaging, and then of course the radiation to a lower PSA is a great endpoint. Love that endpoint, by the way.

Michael Morris: So they could have gotten fluciclovine at baseline, but then they couldn't get fluciclovine as part of their standard of truth for verifying the PyL. But if you look at what people got as their standard imaging as part of their eligibility at their own institutions, 63% of patients got some combination of either a bone scan, a CT or MRI, or in some cases rarely an FDG PET. Twenty percent only got a CTRM or MRI, 10% got either fluciclovine or choline PET as their baseline, and then 7% got just a bone scan as their baseline.

So then when they got their standard of truth imaging after the PyL to verify the PyL findings, 15% had verification of the lesion by biopsy, by pathology. Fifty percent had the verification performed by correlative imaging, such as fluciclovine or choline or a directed MRI. One patient got the radiation therapy to elicit a PSA response. And remember that by design, the focus of this trial was on just positives, right? It was positive predictive value.

So negative lesions didn't count so to speak, a lesion that might be a false negative for example, and if they had no findings on their PyL, which we anticipated about 40% of patients would have no findings, then no standard of truth would be submitted. So that was the breakdown, essentially the comparison of their baseline imaging that they had per their institutional standard versus their standard of truth imaging or pathology.

Oliver Sartor: Well, I thought it was really interesting that 64% of the patients actually had a change in their management plan. I thought that was the bottom line on this study. I mean, almost two-thirds of the patients ended up doing something different as a consequence of the PSMA PET. And to me, that's hugely impactful and indicates that clinicians are really taking this information to heart.

Michael Morris: Yeah. I think that the change in decision-making is supported by the primary endpoint of the study, which was the veracity, the true positive, and the false positive rate of the clinical trial. So let me talk about that first and the implication on the decision-making. So when you looked at the CLR rate, we had set out as the benchmark for success, a relatively low bar, which was based on prior positive predictive values of imaging studies, such as fluciclovine, which the FDA had already approved. So we had set as a benchmark of success that the lower limit of the 95% confidence interval could not go below 20%.

The CLR that was found in CONDOR was 85 to 87% in each of the readers. So it was well above what would otherwise qualify as the benchmark for success. And that CLR rate really held itself through all of the PSA values. So if you look at the patients with a very low PSA, less than 0.5, the CLR was 73%, 0.5 to less than one, 75%, and anything higher than one was 83% to 96% at its highest. So I think that you can say that from a confidence standpoint, what you see if you see a positive lesion, it's quite likely that that is in fact, a true positive lesion.

So that brings us then to the clinical decision-making, the clinicians who were using these scans as part of the patient management for the biochemically relapsed patient. As Oliver just said, 64% of the patients had a change in intended management. And I think what's most important is that for 20% of those patients, that change in management changed from a non-curative goal of care to a curative goal of care i.e. from non-curative systemic therapy to an attempt at salvage local therapy.

And for 30% of the patient population, the change in management went from just salvage local therapy to local therapy that was supplemented by or replaced by systemic therapy. So there were some really important clinical decisions that were being made for these patients. But you can say that those decisions were supported by the performance characteristics of the tracer as demonstrated by the correct localization rate.

Oliver Sartor: Yeah. One brief comment and I think this is for the audience. When I was going through the data to the best of my ability, only about 36% of the patients had a positive scan when the PSA was 0.5 or less. So even though there seemed to be veracity, the actual positive rate was 36%. And for less than one, it was about 50%. So I think we do have to adjust some expectations. You're not going to find lesions on all of these patients, particularly those with a low PSA.

Michael Morris: I think that also speaks to the rigor of this prospectively defined study. So if you look at data from other trials and other publications when you look at those very low PSA values, those patients frequently were collected, not on the basis of specifically having negative standard scans, but they were just prospectively imaged and it happened that those patients had low PSAs. But this patient population is especially rigorous because the patients had already been imaged and were found to be negative and their PSAs had to be in order to be eligible only 0.2 if they had had a prostatectomy.

And 85% of these patients had prostatectomies either alone or with radiation therapy. So we're truly talking about a patient population that's really been screened for having no evidence of disease by standard means. And that's why I think at these lower PSA levels of less than 0.5, you're getting 36% positivity. But I think that you're right, Oliver, that you should go into like your pretest positive probability of ordering a scan with that expectation in mind. Although if the PSA is slumped somewhat higher than it is 51%, 0.5 to less than one in terms of positivity.

Alicia Morgans: So to both of you, actually, I think that many people may be wondering how does PyL and PyL PET really fit into the theme that we have where we have axumin at least in the US approved. We see that PSMA PETs may be coming down the pipeline in the very near future. Where does this fit in and how does this study really help to characterize its best use?

Oliver Sartor: Yeah. Well, Alicia, there was a direct comparison study performed at UCLA to look at the fluciclovine in comparison to the PSMA. And of course, it's in their hands and by the way, the PSMA tracer was actually a Gallium-68 of the different platform than the PyL that's issued here that we use in the PSMA-11. But having said that, there was really no doubt that the lower PSAs that the PSMA PET was more sensitive. And that has been pretty much agreed upon across the board from other evidence as well. So I think the PSMA PET is just a better scan.

We also don't have a lot of the metastasis directed therapy following the fluciclovine scans and that's a big hole in that data. Here, even though there weren't a lot of patients, Mike, there's increasing data for PSMA PET directing metastasis directed therapy successfully, at least as measured by PSA decline. And even with a small randomized trial at Hopkins that would indicate that time to progression was improved. So PSMA PET I think it's going to be a better product than the fluciclovine.

Michael Morris: Yeah. I agree with everything that Oliver just said. I think that the importance in the upcoming year of where we stand in terms of PSMA PET is that there's now evidence for both Gallium and for F 18 PyL that we hope that the FDA will review the evidence for both of these and approve them so that this tool becomes available to practitioners in the US. The reimbursement issues are a separate issue that would obviously need to be defined after that, but it will at least deliver to the clinicians' hands and for the patients, a tool for assessing where your disease is, really that is superior to any other tool from at least the point of view of many prostate cancers, especially in the low PSA range.

Do we know what the best treatments are in terms of reacting to those and how to best deal with these otherwise undetectable lesions? Well, no, not yet, but that's something that clearly now is the next stage of defining the optimal treatment paradigms in terms of how to deal with these lesions that previously were undetected. But I think it's really important to have this tool available to bring our disease detection and then our subsequent treatments to the next level. There's just so much unseen disease that we've in the past had to make decisions about without knowing exactly where the patient's disease has been.

Oliver Sartor: Yeah, no doubt about it. This is game-changing technology and I've done a little bit of estimation in my own mind looking at PSAs and the readouts that you might get from CTs and MRIs. And this is probably at least 10 times more sensitive than the current cross-sectional imaging technology and maybe even as much as 50 times. So it really is a dramatic advance compared to the conventional imaging we've used for years.

Michael Morris: I mean, if you look at CONDOR alone, right? A 100% of the patients had a negative or equivocal findings on their standard scans. Sixty percent of these patients had a positive finding on the PyL scan and 70% of those patients had a treatment decision that was changed by virtue of that scan. So I think that's a really impressive set of tools and actionable information by virtue of the PSMA scan that just didn't exist before.

I think the other important thing is that when you look at the data of both PyL through OSPREY and CONDOR, or the Gallium-68 dataset, their applications really throughout the disease course, there's a pre-op application. There is a biochemical relapse application. There are even applications for patients with metastatic disease in terms of response assessments. So this set of tools really is useful across the spectrum of prostate cancer, even beyond the issues that CONDOR specifically addressed.

Alicia Morgans: Great. So, Michael, just another quick question before we wrap up. And I'm not sure whether the study is able to do this because it certainly met its primary endpoint, but are you and the team following these patients, particularly those who have had management changes to see where their outcomes ultimately stand?

Michael Morris: So that's a very different kind of trial that CONDOR was designed to be. This is really the diagnostic performance of PyL that CONDOR was designed to be in order to satisfy a very specific regulatory requirement. A kind of trial that you're alluding to would really be an outcomes-based study based on a set of interventions. What happened to these patients after they had a treatment change, or even did that treatment change yield a superior result than had the patient not had a change in treatment? That study needs to be done, but that's just not this study, but those trials will clearly need to be done so that we can as practitioners develop the best treatment paradigms for our patients.

Alicia Morgans: Well, this is truly exciting work and we sincerely appreciate that you've taken the time to share the study and a little bit of the behind the scenes perspective with everybody. Thank you both Oliver and Michael for taking the time today.

Michael Morris: Thank you, Alicia.

Oliver Sartor: Thanks, Alicia.