Trials Evaluating the Impact of PSMA PET on Management and Outcomes “Presentation” - Amar Kishan

February 12, 2024

At the 2024 UCSF-UCLA PSMA Conference, Amar Kishan highlights the significant impact of PSMA PET imaging on clinical trials for localized prostate cancer, particularly its role in refining intermediate clinical endpoints like metastasis-free survival. Dr. Kishan discusses the potential of PSMA PET to alter trial outcomes by early detection of distant metastases, referencing trials like STARPORT, INDICATE, and THUNDER to illustrate its influence on treatment decisions and the necessity for direct evaluation of its benefits in clinical settings.

UCLAUCSF video logo stroke


Amar U. Kishan, MD, Professor, Vice Chair of Clinical and Translational Research, Department of Radiation Oncology, UCLA, Los Angeles, CA

Read the Full Video Transcript

Amar Kishan: All right, thank you very much. So I'll be talking about trials evaluating the impact of PSMA PET on management and outcomes. And we've heard a little bit about this, so I'll try to focus more on the localized setting here. So, first, implications on PSMA for clinical trial design, and a brief overview of large trials incorporating PSMA, and then some trials evaluating PSMA. So, at least in localized prostate cancer, it has a very long natural history. So there's been a large push to look at intermediate clinical endpoints or ICEs. The ICECaP group, a number of years ago, found that metastasis-free survival based on conventional imaging, was a good surrogate endpoint for overall survival. Which means, strictly speaking, if you look at that graph there on the Y-axis, is the treatment effect expressed in hazard ratio for OS. On the X-axis is the treatment effect on MFS, and hazard ratio.

There's a very strong correlation, the R squared is 0.92, and it turns out that if something improves MFS by 12%, so the hazard ratio is 0.88, it would have a non-zero impact on overall survival, and it'd be an earlier endpoint. And the main implication for that, in this somewhat complicated figure, is that you can design a trial with an MFS endpoint and it will have an earlier readout, which is very beneficial because these are big trials, cost a lot of money to run, and you want to know your result earlier. So, for example, if the hazard ratio for OS is less than 0.7, which is a pretty big effect, MFS is the preferred endpoint for the trial. For instance, if you look at a thousand-patient trial, so that'd be the yellow curves there, the dashed would be with an OS endpoint, and the solid is with an MFS endpoint. For an MFS hazard ratio of 0.6, you get your readout about a year earlier, which is beneficial.

But how does PSMA factor in? So, due to the increased sensitivity and specificity of PSMA, which we've been hearing about in both the de novo and recurrent setting, PSMA PET/CT will lead to the earlier identification of DM events. How could that impact the association between MFS and OS? And are these DM events of the same prognostic importance? So one question you could ask is, well, when we order a PSMA PET in the recurrent setting anyway, it's typically when a biochemical recurrence is diagnosed. So, is biochemical recurrence a surrogate endpoint? And our group looked at that, it was published late last year, BCRFS, which accounts for both BCR and passing with any cause. And time to BCR, TTBCR, which doesn't incorporate any cause mortality. Those are not actually strong surrogates.

You can see actually the correlation coefficients are quite weak. A treatment that impacts BCRFS or TTBCR does not necessarily translate into an overall survival benefit. So, the clinical event that triggers the PSMA PET may not actually itself correlate with overall survival. And there was also a very nice study published late last year in European Neurology, where they had a cohort of men with oligorecurrent metastatic prostate cancer who had undergone next-generation sequencing. 123 patients with advanced molecular imaging, diagnosed metastatic disease, and 172 with conventional imaging. And it turns out there's far more pathogenic mutations and mutational burden in patients with conventionally diagnosed oligometastatic disease. And if you look at the overall survival, again retrospective, it's far better for patients with oligometastatic disease detected on advanced imaging. So the trigger for the PSMA PET, and even perhaps the disease that's diagnosed at the time, might play a role here. It might not be the same thing as conventional imaging.

So, what are some of the challenges in the de novo setting that we are going to run into? And this was nicely summarized in a paper in JCO. Many of you are probably on it. Know Dr. Hope was on this paper. So, stage migration, changes to eligibility in trials. For example, if you have a trial looking at high-risk prostate cancer, but then you do a PSMA PET and you identify patients with nodal disease, they would be excluded from that trial. More accurate staging, which could be beneficial, inaccurate upstaging, rare, but possibly a patient could have a "rib met", that's identified that pulls them out of the study. And then this could impact the power calculation. If our historical MFS rates are based on patients who maybe had extraprostatic disease at presentation, and we filter those patients out, our event rate's going to be a lot lower. You need more patients to enroll in that trial. So that's a big challenge.

What about the impact on treatment? We heard about this as well, so escalation or de-escalation without proven evidence, even in the localized setting as well as in the advanced setting that Dr. Morris just went over. Lack of adherence to guidelines based on conventional imaging, and using PET for surveillance that could lead to discontinuation of effective therapy earlier than we otherwise would discontinue it. In the BCR setting, all of those caveats apply. And then also, this was alluded to a little bit earlier, unproven omission of local radiotherapy when the PET is negative, or if it shows regional distant disease, or incomplete understanding of PSMA PET in the radio recurrent setting. Maybe these patients don't have actual disease, there could be false positives in that setting in the prostate, or maybe these patients are being over-treated with ADT or local therapy. So this can be very complicated.

So, those challenges basically are that the surrogacy of IMFS, or based on PSMA MFS, is likely to be weaker than traditional MFS, which might impact trial design. And the widespread uptake of PSMA PET will impact accrual into trials and also impact results. And these are ongoing trials. So, it will be an issue for basically most of the trials that are ongoing currently. How are larger trials that are more recently designed incorporating PSMA? And obviously, this is only a brief overview. The STARPORT trial, which was mentioned in passing before, a very important trial looking at patients with either de novo or recurrent oligometastatic prostate cancer, defined as one to 10 lesions, including PSMA PET as identifying that in the standard of care PET CT that's listed there. Those are randomized to receive either standard of care, systemic therapy, or systemic therapy, plus PET-directed local therapy.

Now, the primary endpoint is castrate-resistant prostate cancer-free survival. But the secondary endpoint, rPFS, actually does incorporate PSMA PET. So you can actually ... The second bullet point there, radiographic progression includes unequivocal new PET-positive lesions and/or unequivocal PET progression. And that endpoint will be stratified based on conventional imaging versus PET-based imaging. This trial was very recently revised actually to include de novo disease, because of the approval of PSMA PET, and expanded to go up to 10 lesions because of the widespread adoption of PSMA PET. It was originally only a recurrent disease with one to five lesions. The INDICATE trial, which is an ongoing ECOG trial for patients with biochemical recurrence after prostatectomy, if it's more than one year later, PSA 0.4 or higher, if it's within one year 0.2 or higher, all of these patients get a standard of care PET, which is either fluciclovine or PSMA.

And then they're actually randomized after stratification for the results. So if it's a negative PET, the randomization is adding apalutamide to Lupron plus salvage radiation, or just standard. If the PET is positive for extraprostatic disease, then all patients get standard radiation plus apalutamide and Lupron. And the randomization is adding metastasis-directed therapy. Here, however, the primary endpoint is PFS, which is based on conventional imaging progression only. So it doesn't incorporate the PET into the endpoint. It's seeing whether using this information upfront may help improve historical control outcomes. But a secondary endpoint that's prespecified does include PET-based progression after rising PSA. There is an interesting exploratory imaging endpoint, but it's only really baked in and mandated for fluciclovine, not for PSMA PET. And then we have the THUNDER trial, which was recently promoted on X by Piet Ost. So this is just opening. It's for localized high-risk prostate cancer, and possibly clinically known positive disease.

All patients get a PSMA PET and Decipher. The patients that have a negative PET and low Decipher are enrolled in a phase two portion, where they get radiation plus darolutamide, instead of what they would standardly get, which is an LHRH agonist. If they have a positive PET or a high Decipher score, they're randomized to the addition of darolutamide to an LHRH agonist. So the primary objective of the phase two portion is quality of life, for the phase three portion is actually ppMFS, which is looking at PSMA PET progression to MFS. So, as you can see there, the tertiary objective is actually to establish PSMA PET-CT as a re-staging tool for MFS.

And finally, an overview of large trials evaluating PSMA PET. Just a shout-out to the EMPIRE-1 trial, which obviously did not look at PSMA PET, but was a very prominent study looking at the incorporation of fluciclovine for radiation planning.

Dr. Hope was mentioning at the beginning, it's rare to see some of these papers making it to Lancet. This one was in the main Lancet. So, looking at, again, patients for salvage radiation, whether they get a fluciclovine that can influence plan design or not. And essentially what it found was that at three years, median event-free survival was significantly improved with the incorporation of fluciclovine before salvage radiation was delivered. 76% versus 63%. UCLA, Dr. Calais, Dr. Nickols, led the PSMA-SRT trial, which is of a similar design as patients that are presenting for salvage radiation after radical prostatectomy, randomized in a one to 1.3 fashion, to a control arm standard of care imaging, which could include fluciclovine PET versus PSMA, and then using that information to plan radiation. The primary outcome was event-free survival at five years. The trial enrolled from 2018 to 2020. It is closed. There's a poster presentation, which I encourage you to see by Wes Armstrong, a second-year medical student at UCLA, focusing on management changes, which is a secondary endpoint.

Plans were available for 92% of these patients. And fluciclovine was used in 43% of the control arm. And what we found was that there were major changes in radiation plans detected. 45% in the PSMA PET arm versus 22% in the control arm. I think it's very important to have that background rate, because obviously, there have been single-arm studies, so that plans change. But here we're taking into account that many things can change. The floor, that what would change if you didn't have PET is not 0%. We get other scans, an MRI, you can get a fluciclovine, maybe a Decipher test. Other things can influence this. So it's important to know what's the delta that the PSMA is providing. And it is pretty substantial. Treatment escalation was seen in 29% versus 12% of patients. It's a nice figure. I don't want to steal too much of the thunder from the poster presentation, but you can see as more advanced disease is detected, there's significant treatment escalation, and perhaps even de-escalation if there's no disease found on the PSMA PET.

The DRT trial, Dr. Nickols mentioned in passing as well, was a multicenter randomized trial for men with unfavorable intermediate risk or high-risk prostate cancer, attempting to look at how PSMA PET-CT might influence five-year progression-free survival. Open for a year, very difficult to enroll because there were competing trials of PSMA. And then once the FDA approved PSMA, there was no chance that this could enroll. So we closed it early, didn't enroll 54 patients and have forty-six patients. A manuscript looking at this is under review.

Mentioned in passing, the PATRON trial from Canada that's looking at patients with a high risk of treatment failure, either untreated clinical node positive or high CAPRA score or biochemical recurrence. Everyone gets a conventional imaging-based treatment plan, and then randomizing to PSMA PET guidance, or just treatment based on the conventional imaging plan. Then point of failure-free survival, a very large study, 776 patients. They have a very rigorous way in which the PET information can be used to guide treatment.

And then you have the EMPIRE-2 trial, which is ongoing, and is essentially comparing fluciclovine-based salvage radiation planning against PSMA PET-based planning. So in summary, several large trials have incorporated PSMA PET-CT into their design and will indirectly provide invaluable information about the utility of PSMA PET. And there are several trials underway that are directly evaluating the benefit of PSMA PET, particularly in the salvage RT setting. Thank you.