Limitations of PET Imaging in Prostate Cancer - Stefano Fanti

August 29, 2020

Stefano Fanti joins Phillip Koo in conversation on establishing the PSMA PET/CT, as the gold standard with regards to imaging for prostate cancer, expectations, accuracy when it comes to imaging, how that CT should be performed, and limitations with regards to trial design for diagnostic imaging agents. Dr. Fanti offers advice for nuclear medicine physicians and radiologists in the community with regards to the multidisciplinary care for prostate cancer.


Stefano Fanti, Professor, Department of Experimental, Diagnostic and Specialty Medicine - DIMES at the University of Bologna. He is Director of Nuclear Medicine Division of the PET Unit at the Policlinico S. Orsola and Director of Speciality School of Nuclear Medicine at the University of Bologna.

Phillip J. Koo, MD, FACS, Division Chief of Diagnostic Imaging at the Banner MD Anderson Cancer Center, Gilbert, Arizona, USA

Read the Full Video Transcript

Phillip Koo: Welcome to APCCC 2019 here in beautiful Basel, Switzerland. Today, we're very fortunate to have with us a long-time friend, Dr. Stefano Fanti, who's the Director of Nuclear Medicine at the University of Bologna. Welcome, Dr. Fanti.

Stefano Fanti: Thank you. My pleasure to be here.

Phillip Koo: We're very fortunate, and just congratulations on a wonderful lecture today. Really establishing the PSMA PET/CT, as the gold standard with regards to the best performing imaging agent that we have in prostate cancer. I think there are expectations when it comes to imaging that some people have, that imaging could somehow be perfect in all instances. Can you talk a little bit about the limitations of PET with regard to imaging?

Phillip Koo: Yeah. Sure. Well, first of all, limitation is, you mentioned that gold standard. Indeed, it's an optimistic vision. I agree on that. But, I'm biased at being a nuclear medicine physician. So gold standard means that you have to prove, beyond any doubt the fact that what you see is really disease. And unfortunately, we cannot really state that that's been done, because it's the pathological proof of any findings as not extensively been done in the literature. So the first limitation is that PSMA is not an absolute, specific prostate cancer tracer. So you can have hotspots for other reasons. Many of them can be ruled out with some experience in reading the images and understanding the patient history, but others are more tricky.

It's important to mention that the number of those potential false positive findings, is very low. So it's in the range of five to 10% depending on many [inaudible 00:01:36] of the patients. But we have to remind that it's not equivalent hotspot to metastasis for the patients. And the other way around, the limitation on false negatives is mostly related with our capability to see the tumor. So we need a certain tumor burden, a certain dimension of the lesion. So, of course, we can never see micrometastases of few cells in a one millimeter lymph nodes. That's far beyond our capability.

Phillip Koo: Sure.

Stefano Fanti: But nonetheless, I'm very happy that you catch the message. It's the best available imaging method in several scenarios, including biochemical recurrence, including planets of a salvage radiation therapy, including staging of high gross patients. So it's the best approach so far.

Phillip Koo: So in your opinion, what is the resolution at which you feel that PET/CT, PSMA PET, isn't as accurate?

Stefano Fanti: Well, we have to distinguish between the special resolution. That's to say the dimension to which CT and PET can go in term physical, that's around five, four millimeters. Of course, CT can go much better, but it's the PET components that cannot resolve anything lower than four or five millimeters. And then we have a biological resolution, let's just say the burden of the tumor, we need a certain amount of receptor to be overexpressed by a large amount of cells. So definitely micro matter. If we define something like 1000 cancer cells, we will never be able to see that with PET so far.

Phillip Koo: So Dr. Davis mentioned about the importance of the CT part in a PET/CT, and how there's a lot of variation in how that CT is actually performed. Can you talk about, from your experience, how that CT should be performed?

Stefano Fanti: Yeah, that's a very good point. I mean usually, almost nobody is carrying out PET scan alone. It's always PET/CT and there's people doing a diagnostic CT, so with media contrast, early delayed image. So a fully diagnostic CT that has to be reported by a licensed, said radiologist for example, in Europe where the specialty are completely divided and that's one point. It could add information, it will add a little bit of radiation exposure, it will add time, it will add cost. Other centers are not doing that because they only use a low dose CT, just for anatomic correlation and to have some gross information. This is done by everybody. Of course, it's mandatory.

And I guess that what I really liked by Dr. Davis' lecture is that he mentioned it very properly that CT is a fundamental part of that. It's not just to say, okay, it's a lymph node rather than a bone. But for example, if you find a suspect hotspot, within the lungs, you have to look at the characteristic, CT characteristic of the nodule because it could be a met, but it could be also a different primary tumor. Again PSMA is not absolutely process specific. We have seen several examples of a second adenocarcinoma of the lung, identified by that. And in many cases, we can have information about the CT appearance, something other than itself. And that's the same, I mean, a ganglion can be easily seen on CT and you will avoid to call it, as a false positive finding. If you do identify that in the CT, you cannot do that in the PET one.

Phillip Koo: Great. So a lot of the discussion today regarding PSMA imaging is the real outcry for more outcomes data, and just to understand what the clinical impact is on patient care after management changes occur. So can you talk about some of the limitations with regards to trial design for a lot of these diagnostic imaging agents?

Stefano Fanti: Yeah, that's a great question. Probably the more challenging question to the practice of nuclear medicine physicians, that's to say we are very used to trial design, and in order to demonstrate the superiority or at least the non-inferiority of a drug, over another existing best treatment available. That's clearly known by everybody and led them to the registration of the pills. Okay. But that's completely different for imaging. First of all, we have one bias of not being directly at therapy. So you put two different variables, that's to say imaging show something and then you have decided to auto treat. So you have two different variables that do interfere each other. So the trials should be much larger, much more expensive. And it's not that easy to establish which is the gold standard, of the therapy in any situation. Because we will face situation that we haven't seen before. It's like that typically oligometastatic disease as identified with PSMA.

And the other clear point is that there are no big pharma company beyond it because several PET trials should have a very small market and some of them are not patented at all. So while all the big pharma knows very well how to run, and especially to brand, and to fund, and to pay for the very high cost of a trial, there's nobody doing that. So luckily enough in Australia, for example, Movember has been, as a foundation, has been funded many on the ongoing prospective trial, it started enrolling thousands of patients. This is something that has not been possible to be done in the rest of the world so far. And there are several other complications as I say, how you interpret the data from the images. I mean, if you're doing a pill, you give it or you don't give it. But as you mentioned it before, they may be tricky finding on imaging. So even reporting, as an interpretation should be incorporated in part of the trial and that makes things even more complicated.

So, okay, we should really work better. And of course in close conjunction with the clinicians, because we only run the trials to demonstrate the accuracy. That's to say our images are better than conventional images, that's not enough. You have to demonstrate that it led to a change in patient management and this change in patient management led to a beneficial, hopefully in overall survival.

Phillip Koo: So maybe this is a good call to action to the pharmaceutical companies to step up and assist and support these trials because I think these types of trials will have a huge impact on the clinical care of our patients.

So you did mention sort of briefly about theranostics, and clearly nuclear medicine is getting into this new arena of treating patients while also diagnosing disease. What advice do you have for nuclear medicine physicians and radiologists in the community with regards to the multidisciplinary prostate cancer team?

Stefano Fanti: Well, right now we are incorporated to a full extent in multidisciplinary team, that's occurrent. I mean, at the very beginning when we only can provide information from bone scanning, it was not realistic to have them around because the information was really minimal. With PSMA scanning, we have been, and even before with colon scanning some years ago, we already started to have a more fruitful collaboration. And with theranostics we are definitely part, it's just been on paper recently from the Peter Mac group in Australia, about the fact that they've been finally fully recognized into the MDT. Well, it's a recognition. It's approved that nuclear medicine is growing very rapidly and making the real game scenario, finally.

Phillip Koo: Well, thank you very much for joining us today. Again, wonderful talk and we look forward-

Stefano Fanti: Thank you, thank you very much.

Phillip Koo: ... to hearing from you again soon.

Stefano Fanti: Thank you very much. My pleasure. Thank you.

Phillip Koo: Thanks.