Examining Differences Between PSMA PET Imaging Agents "Presentation" - Wolfgang Fendler

February 5, 2024

At the 2024 UCSF-UCLA PSMA Conference, Wolfgang Fendler discusses the relevance of different PSMA PET agents in clinical practice, emphasizing that while there are various ligands with kidney-dominant or liver-dominant excretion, their clinical impact appears minimal, suggesting PSMA-PET is a unified class of imaging with different logistical aspects. Dr. Fendler highlights the importance of understanding each ligand's physiological distribution and common interpretation pitfalls, advocating for inclusive clinical trials that accommodate various radioligands for broader research and clinical application.

UCLAUCSF video logo stroke


Wolfgang Fendler, MD, Department of Nuclear Medicine, University of Duisburg-Essen and German Cancer Consortium (DKTK), University Hospital Essen, Essen, Germany

Read the Full Video Transcript

Wolfgang Fendler: Set for the second presentation. While the slides are coming up, my name is Wolfgang Fendler. I'm a nuclear medicine physician from Essen University and before starting with my talk, I, of course, want to thank all the organizers, especially Johannes, Tom, and Jeremie, for inviting me to this focused conference on PSMA imaging and treatment. My topic today is, do the differences in PET agents actually matter?

Starting with some disclosures on my side, and then going right in with the different ligands. If you look at the publications of recent years, you find these MIP bio-distributions and very different ligands that have been published. I just took one of the images from our guideline to separate these ligands into kidney-dominant excretion and liver-dominant excretion. Of course, this is not a complete overview. There are many more ligands than these.

They are gallium-68-labeled, F-18-labeled, zirconium, copper-64, which then leads to the questions, how many radioligands do we actually need for our clinical practice in the future and is it not possible that these radioligands are actually one class of PSMA imaging? Then we have a little bit of different flavors in terms of logistics, but all these ligands do the trick in terms of our clinical questions that we want to solve on the patients.

I want to go a little deeper in terms of clinical evidence and also clinical experience on the differences between those ligands. These are the three levels I would like to look at. Three levels are logistics, so looking into the radiopharmacy production availability, facts, which is basically clinical evidence. Do we have studies out there comparing these different ligands? Is there a difference in the studies? And then also experience about our clinical experience, clinical application of these radioligands, and maybe clinical pitfalls that we have encountered in our daily routine.

Optimizing the radiopharmacy costs. I just want to give you quick insight into the Essen experience and how we use different radioligands and how this could have helped to optimize the radiopharmacy logistics. Until 2021 we used gallium-68-PSMA11 labeling, which means that we had up to three daily PSMA11 productions. This would be 15 per week. Depending on the age of the gallium generator, this would be an elution every three hours and then leading to nine patient doses per day, ideally. Usually, it's a little bit less, but ideally nine patient doses per day, 45 patient doses per week. There's some uncertainty in terms of the age of the generator. There's some uncertainty for patients not showing up or showing up later. So usually on a daily basis, you would scan fewer than nine patients, but ideally nine. Of course, you have a short half-life and you have a lot of shipments for gallium-68 because of the short half-life, and you need to ship all these synthesis to the different PET devices.

So in total, about 60 full-time employee hours for 45 doses, and of course in 2021, we figured this is too much. We need to optimize and switch to F-18 compounds. Regardless of which compound you're producing, we switched to Gallium-PSMA1007 for our own in-house production that's cyclotron-based. There are two productions per week. There's a stable yield of 60 to 80 gigabecquerels per batch. This corresponds to about 30 patient doses per day, so 60 per week. There are more flexible scan hours based on the two-hour half-life, and of course, there's a much lower shipment rate of about six dose shipments per week.

So in total, 22 FTE hours for 60 doses. In terms of logistics and applicability in terms of your staff hours, the F-18 clearly wins in German efficacy, and we're all here, so that's why we decided to switch also on a daily routine to the F-18 because we just have a more stable yield and better logistics and better efficacy. Of course, there are clinical data I promised we will look into the facts. Clinical data on clinical trials comparing those different ligands really head-to-head in high quality. This is a prospective trial that compared PSMA-11 and PSMA1007 head-to-head for initial staging of prostate cancer, and this clinical trial enrolled 60 patients. Both scans were applied to the patients, and they found almost perfect concordance for the primary staging for both different ligands in the 60 patients with a kappa of about 0.19 or even higher.

The dominant intraprostatic lesion was similarly detected by both radiotracers, and there are some patients like this patient example shown here where the PSMA1007, due to the low urinary excretion, showed additional local lesions or secondary foci in the local prostate as shown here by the two arrows. So some mild advantage, but probably not so clinically relevant in those patients because all the lesions have been detected. Let's look at other head-to-head comparison trials.

This is a study with a mixed cohort from initial staging, restaging, and metastatic disease restaging patients comparing again 1007 to PSMA-11 head-to-head in the patients, and there was a higher rate of concordance in 92 patients. Both ligands did accurate staging, and PSMA1007 even delivered more equivocal results especially for bone lesions, which also some of these were subsequently confirmed as prostate cancer. So more equivocal results as compared to PSMA-11, and here's one of the cases of a mismatch highlighted in a patient where PSMA-11 even showed adrenal metastasis due to the low liver uptake that was not so easily seen on the PSMA1007 scan due to high liver uptake, but in general, a high concordance rate in this trial comparing head-to-head.

Another head-to-head comparison trial, again prospective with a very thorough design, enrolled 82 patients and they were randomized to two tracks, and basically in the two tracks, patients received either first Gallium-PSMA-11, followed by F-18 PSMA-11 or just the other way around. So very high quality level for a head-to-head comparison. And the time difference was a maximum of three weeks between the two scans. So it's really the same setting for those two scans, and again, the prospective trial found that the F-18 compound is non-inferior to PSMA-11 or you could also rephrase it and say it delivers equal accuracy, and the primary endpoint was met. If you look at those positive predictive values and detection rates, you see they are nearly equal among all the readers that were included in this prospective trial which was published in European Urology.

Of course, if you take all the evidence together, you can now look into a meta-analysis of differences of ligands. There's one systematic meta-analysis published here in Nature Prostate Cancer and Prostatic Diseases. On the right side, you can see the forest plot on 24 studies comparing those different ligands head-to-head either in the BCR setting or in the primary staging setting, and those bars are either shifted towards the F-18 SUV or the gallium SUV depending on which of the SUVs have been higher, and I have to say I don't see a major difference for the different ligands in terms of the uptake and the total uptake on those lesions. Also, this meta-analysis found that when comparing DCFPyL and Gallium-PSMA-11, there have been similar lesion detection rates. When comparing PSMA1007 to the other two, PSMA1007 showed a greater local detection rate in these head-to-head comparison trials.

But if PSMA-11 was applied together with furosemide or protocols that decrease the urinary excretion rate, then again the local lesion detection rate was quite similar and similarly higher for 1007, and of course, 1007 showed equivocal findings, especially for bone due to unspecific bone uptake. This is one of the examples that we encountered in our daily clinical routine, and now I'm going to the clinical experience part on 1007. The F-18 ligands 1007 and rhPSMA can show unspecific bone uptake, as in a patient with biochemical recurrence, a PSA level of less than four. We scanned the patient with 1007 and found focal uptake in the rib with an SUV of five or higher, and to double-check, the patient underwent subsequently within two weeks PSMA-11 and also underwent a bone scan, and both these scans we could not confirm bone uptake. In the lower row, you can see a patient with scapular focal uptake, and again on PSMA-11 and bone scan in the same patient performed to evaluate those lesions.

There was no uptake and subsequently, it was not confirmed that there was bone metastasis. So, these are typical examples of unspecific bone uptake on 1007, and to also show one case on rhPSMA, I borrowed a slide from Matthias Eiber from the Technical University who performed a lot of rhPSMA scans, and this patient with biochemical recurrence and a PSA level of 0.8. The rhPSMA was able to demonstrate local recurrence, local uptake, but also showed some unspecific uptake in the spine due to active or degenerative changes here, but again, it shows focal uptake that should not be mistaken for bone metastasis. So, both rhPSMA and PSMA1007 should be evaluated very carefully when it comes to bone uptake. There's a pattern on where these bone uptakes show up. Most of these are in ribs. So, 41% have been found in this trial from our site in ribs.

But of course, there are other sites which are much more challenging. If you look at pelvic bones or vertebrae or scapula, it's not quite clear just by the location of the lesion to make a judgment whether this should be unspecific uptake or not, and especially these locations that are typical for prostate cancer, pelvis, and also spine, are much more challenging to differentiate between unspecific uptake and true metastasis using those F-18 PSMA ligands. We looked into our database and tried to find out if the 1007 or the F-18 PET makes a difference in terms of the rate of diagnosis of bone metastasis and we did not find a significant difference. So, the rate of bone metastasis did not change from the transition from Gallium-PSMA-11 to 1007. It's not a head-to-head comparison but a large database assessment, and we believe this is due to adjusting your clinical experience in how to analyze and rate these lesions.

And this is our recommendation from clinical practice on how to avoid false-positive interpretation for 1007 or rhPSMA: for the ribs, of course, avoid oversensitive interpretation, be very skeptical about focal rib uptake without any other correlate, then take into account the CT part if you see any sclerotic or lytic lesion, and of course, consider the PSA level if it's justifiable or realistic in a patient to have distant bone metastasis if there's just one foci, for example, in the ribs that light up on these scans. How to handle different PSMA radioligands in clinical trials. I just wanted to show two examples for this. The PROTEUS study from Janssen included PSMA-11 quite actively. Patients undergo PSMA-PET at three months post-adjuvant treatment at biochemical failure, and then every six months until then. This is a randomized trial where patients with initial high risk are being randomized to either neoadjuvant apalutamide and then followed by local treatment then followed by the continuation of apalutamide or placebo.

And these patients undergo imaging very actively to assess for recurrence and for their disease status, and this trial handled it in a way to be very inclusive. So all these different ligands are allowed within the prospective clinical trial, but there's no switch in ligand allowed between the different radioligands, which I find reasonable to avoid any difference in uptake or sensitivity or biodistribution. The second trial is the PRIMORDIUM study also from Janssen, including patients that have more than one local regional lesion and a high-risk biochemical recurrence, and also these patients undergo regular PSMA-PET scans, at screenings, six months, 12 months, and then regularly thereafter, and this trial assesses then the outcome under these two different treatment schemes. So radiation therapy plus hormonal therapy and again, the trial does the same. All these ligands can be included and they are all allowed, but they should not be switched within the trial to avoid any difference in biodistribution for these different ligands.

Looking specifically into the topic of eligibility for PSMA therapy, I just want to highlight there's an update published in JCO supported by ASCO that clearly states now that all these available ligands, PSMA-11, PYL, and also rhPSMA, can be used as radiotracers to determine eligibility currently for PSMA treatment, and I find this is a reasonable statement for the clinical routine and also makes sense to increase availability for the future. How to assess eligibility for PSMA therapy, and this is based on the VISION criteria bound to an uptake of your tumor lesions higher than liver for the kidney-dominant excretion lesions. All these mentioned here under liver reference, the rhPSMA, PYL, PSMA-11, but there's one exception for patients undergoing PSMA1007. So, more European focus. There should be a spleen reference due to the high liver uptake in PSMA1007 for these patients, and it really makes a difference.

Here's a case example of a patient with bone metastasis on PSMA1007 and taking the spleen reference, this patient would be eligible for PSMA therapy and the patient was enrolled, underwent four cycles, and showed a good biochemical response. Based on a liver reference, this patient would not have been eligible for PSMA therapy. So, I again find it reasonable to switch to a spleen reference for PSMA1007 for the assessment of eligibility. So, I want to summarize our view on different ligands. Does it matter? Does it make a difference to use the different radioligands out there currently available? I would summarize that the clinical evidence does not support a clear relevant clinical impact of the radioligand choice of the ones that are realistically available currently. I find PSMA-PET is one class of imaging. However, there are, of course, specifics that we need to take into account when making an interpretation, which is especially the urine and bladder uptake for those with kidney-dominant excretion, PSMA-11, and PYL, and of course, bone uptake and potential unspecific bone uptake for rhPSMA and PSMA1007.

Therefore, knowledge of physiologic distribution and common pitfalls is very critical for PSMA-PET interpretation. We need to know which ligand we use. We need to be trained for the ligands. We need to have a good experience, and then if we apply it, we see for those with low urinary excretion, 1007, we see a higher local detection rate. However, rhPSMA and PSMA1007 are known to show bone foci, bone uptake, and these bone uptakes, especially in the ribs, need careful assessment to avoid equivocal findings or even false-positive assessments using those ligands. And of course, I want to make a statement for future development. I find that clinical trials should be inclusive of those different radioligands and should be flexible, and we should really support the use of different radioligands also in research for the future. With this, I thank you very much for your attention, and I'm very much looking forward to the next talk by Courtney on pitfalls. Thank you very much.