Jeremie Calais: Hello everyone. Hello, the PSMA-PET Academy program. My name is Jeremie Calais, I'm a Nuclear Medicine Physician. I work at UCLA and I lead the Clinical Theranostics Program of the UCLA Nuclear Medicine Division. Today, it is my great pleasure and honor to share my experience with PSMA-PET. I'm going to talk about understanding the physiological PSMA biodistribution and what we can consider as pathological uptake on the scan. Before I start, just to be systematic, put my disclosures here. I will pause a couple of seconds. And now we're going to start.

So just before, I just want to make sure that we all talk about the same thing and everyone is on the same page. PSMA is a biological target for prostate cancer PET imaging, and one among many others. There are many others. You can do the metabolic tracer with choline, fluciclovine, FDG, bombesin. And today we're going to talk about PSMA. But I want to make sure the audience understand that PET-CT imaging encompass a various potential target for prostate cancer imaging we can do. There is of course, historically the bone microenvironment with bone scan and fluoride PET, but also all the metabolic tracers that include FDG, choline, fluciclovine, and now we have this cell surface protein expression of PSMA that is becoming the current standard of care because that's the best modality and the most performing one. But I just want to make sure that we understand that PET-CT encompasses various of potential tracers targeting different pathway for prostate cancer imaging.

PSMA. So what is it? PSMA is a bit a misleading name. Of course, it is prostate specific membrane antigen, but it is also known as FOLH1 or glutamate carboxypeptidase II or N-acetylaspartylglutamate. The roles we know about this type II membrane glycoprotein enzyme is that in the neurologic system it acts in the neurotransmitter mechanism and in the digestive system it acts on the folate metabolism. Folate is close to vitamin B9 and the hydrolyzation on the enzymatic site, on the external part of the cell, cleave the glutamates into monofolate.

So that are the only roles that are pretty established. I think there are many others and we just don't really know them, such as in the prostate, for example. And why do we use that for prostate cancer imaging? Let's have a look first at the expression in normal organs. When you look at the Human Protein Atlas, you see the main organs here that have some PSMA expression and this correlates very well with the biodistribution you observe on the PET-CT PSMA targeting. You see the lacrimal and the salivary gland, you see the liver, the spleen, the kidney, the tubular proximal cells express a lot the PSMA, and you have the duodenum and often also the bowel that has a significant uptake. It is excreted to the urine. so usually the bladder contains some radioactive urine, as you can see here.

Now, if you look at the PSMA RNA expression in cancer, here you can see different cancer types, and the one that you see here is prostate cancer, and you can see that for some reason, the level of overexpression in prostate cancer is very, very high, up to 100 and 1000 time. And if you are able to target that, you can see that you get clear images with very high contrast, high tumor to background ratio because it translate this very high overexpression.

The intensity of PSMA overexpression as assessed by IHC in prostate cancer carries some prognostic information, and it is related to the aggressiveness of the cancer. You can see here that the more you have PSMA expression and the more intense is the staining, the more aggressive other lesions and it correlates as well with the outcome of patients. These were done on primary tumors with biopsy samplings.

And this overexpression is present in almost all prostate cancers. When you look at a summary of IHC studies here, you can see that almost all lesions were positive somehow for PSMA and pretty intensively. At the end, you end up with almost 95% of the prostate cancer cell that do express the target. So that's why it's very relevant for prostate cancer imaging.

And that's how you get these pretty impressive images with very low background here. And you can see clearly these black dots that correspond to the prostate cancer lesions that are shown here. In that case, that's a patient with a initial PSA of 7.1. These nine bone scan and CT were negative, and you can see clearly, very easily in blue, some pelvic lymph node metastasis, in yellow the primary tumor and in the red bone metastasis. And this really changed the whole staging of prostate cancer.

In the registration trial that were used for FDA approval of PSMA-PET, either for gallium 68, PSMA 11 or flur 18 DCFPyL, I put here the studies, the specificity and the predictive positive value was very high, up to 95% of specificity and 75 to 90% of PPV, which means that when you see some signal in the lesion, it is almost very, almost the time, prostate cancer, but it's not 100%. And now we're going to go over some few pitfalls that can arise and how to interpret this, because although that's one of the most easiest imaging modality you can read now, it is still a medical imaging technique that has some limitation, as always.

So let's start with the prostate gland and the primary tumor. There is a significant data that correlates the intensity of the uptake, as measured by SUV, to the Gleason grade or the Gleason score. You can see here different studies. And what you can see is that there is a significant overlap in the value of SUV between zero and five and even zero and 10, as you can see here. Of course, the higher the SUV is, the more likely you have an aggressive prostate cancer, up to Gleason grade five here. This is in the resident primary trial from Louis Emmett in the European Urology paper. So you can see there is a significant overlap, but if you have an SUV max of above 10, you are pretty much sure and very specific that you are dealing with the clinical significant prostate cancer with a Gleason seven at minimum.

So when you look at that in into images and in patients, when you have focal intense uptake with a very high tumor to background ratio, it is clearly 100% prostate cancer, such as here, easy read.

When you have also a focal uptake, but this time the uptake is very low, you can see here SUV of four or three, and you're not really sure, this is more low probability of clinical significant prostate cancer. Maybe it's a Gleason six lesion, maybe it's very small, but your probability of a very aggressive prostate cancer decrease a lot. When it's diffused, but you still have intense uptake, such as here, you see that just a big prostate cancer that takes a lot of the gland and that's why you have this diffuse uptake, just because with the intensity of 13, it is also a very high probability of prostate cancer. This case was confirmed by histopathology.

But if you have a diffuse uptake and this time a low uptake, then it gets... You get much less confidence in calling this as a positive lesion. For example, here, this kind of diffuse uptake you see here, I would call that a negative or borderline at the end. The PETs imaging information is not very helpful in that case, you are at limitation of the technique or just having a low grade cancer or negative scan.

Let's now look at the PSMA-PET uptake in the lymph nodes. The, the main pitfall is the ganglia. For some reason, the [non-milinic 00:09:44] nerves, they have a little bit of PSMA expression on it, and sometimes it can mimic some lymph nodes and metastasis because it is located at more or less locations where you would expect to see some instant metastasis. It has some uptake in that. You can see that the uptake is usually pretty faint, ranging from 1.5 to three. And so it can be tricky.

The way you saw this out, you look at the shape on the CT. So sometimes you find it where a continuum with the nerve, or you have this typical band shape or tear drop shape, or nodular shape. And the usual best way to get around that is to look at the coronal images, scrolling to the PET on the images and you try to see symmetry. Usually when it's on one nerve, it is also on the other side and that other level from the sacral to coeliac to cervical ganglia. So that's how you saw this out. After a couple of cases, you get rid of this pitfall pretty easily.

Here I show another image where you see clearly in the nerve right away here some uptake, and here it's clearly in the nerve. And you know that... Then you see the uptake that can be seen on such nerves and you have to translate this image to potentially more tricky location close to the vessels.

Also like in any nuclear medicine technique with any radiotracer, you can have some non-specific uptake due to inflammatory process, granulomatosis. And again, here, what you look for is really to have some kind of symmetry between the nodes. You can have that often in the hila nodes, the mediastinal lymph node, and also in the inguinal or distal left external ileac areas. Usually you look for kind of symmetry, multiple sites, faint uptake. It is usually not prostate cancer. You would need other arguments to call that prostate cancer lesions.

 The physiologic activity in the digestive system can be tricky sometimes if it's very close to a tiny lymph node structure here. So make sure you always look at your images in nine view, for example, this tiny focal uptake here that was seen only in axial images was mistaken for many potential metastasis, whereas when you look at the coronal views, clearly it was just digestive activity.

And at the end, it's always the same rule. If it is focal, if you have high tumor to background ratio, the SUV max doesn't matter so much. For example, in that case, you see these very tiny lymph nodes 4 mm only, the SUV max is low, three only because these are tiny lesions so you have some partial volume effect that decreased the intensity of the SUV, but because it is very focal because you have no background around it and you have these high contrast images, it is a very high probability of prostate cancer. In that case, you have three location on both sides, which it increase the confidence to call this pathological uptake.

Same case here, you see this very tiny lymph node on the obturator area on the right side. It is a tiny lymph node. This would have never been caught on a CT before, but because it is clearly separated from the background activity, you can see a bit of urine radioactivity here, but with a clear separation and a clear focal pattern, then you can call that lesion suspicious for prostate cancer metastasis. And this was also confirmed by histopathology.

Now, there are some tricky case when, for example, the SUV's in the two to four range, two to three, and you have some uptake here. These lesions turned out to be negative, even after salvage, pelvic lymph node dissection. And you will have always some kind of overlap between these lesions that have low uptakes. Sometimes it's very difficult to make a call.

I show some other example of very tiny lymph nodes with very faint uptake. These ones are hard to deal with. So you need to put to factor in other parameters to make your final call. Is there other node metastasis already that doesn't change the patient from N0 to N1 or from M0 to M1? Is there aggressiveness features, a high PSA level, things like this and more arguments to at the end make your final call. I will go over that in terms of communication between the referring urologist and the PET-CT reader. I will come back to that later.

Now, let's look at the bone. The bone is a common site of prostate cancer metastasis, of course, and you have multiple types of process that can mimic prostate cancer metastasis if you don't look carefully at the CT. Basically anything can show a little bit of increased uptake as long as you have some bone remodeling, bone inflammation, increased vascularization, all this will create some kind of non-specific PSMA uptake. The PSMA is also present on neovasculature cells. So each time you have some healing process and neovasculature involved, you can have a little bit of PSMA uptake. Here I show some regenerative process, hemangioma, Paget disease, avascular necrosis, but all these, you look at the CT and you get rid of the pitfall right away by looking at the CT.

Here is another example of osteoarthritis with subchondral fracture. The patient have got even a hip replacement so we have full pathology to really call that a negative for prostate cancer. And of course, just based on CT you would not call that suspicious for prostate cancer.

Now, sometime you have more focal uptake and that are associated with lesions in the bone that can have different patterns. One thing is that if you have this kind of lucent lesion with a sclerotic rim, usually this is not aggressive. This is not prostate cancer here. In that case, it was a degenerative cyst, but you can see that the uptake is focal. And if you look at the PET only it can mimic potentially metastasis.

So you look at the CT, kind of similar pattern here with a lucent lesion with the sclerotic rim, this was a cartilage lesion. And this you can have also a little bit of PSMA uptake. So you have some false positives, but just looking at the anatomic imaging modality and CT and MRI, you can sort out that pretty easily.

When it gets tricky, it's for the bone trauma. So here it's an easy case of a vertebral fracture. So you see the diffuse uptake following the fracture and also in the rib here as well. And the ribs, these are the more tricky one. The ribs you can have a little bit of trauma, a little bit of the time. People can curve. People can get a little hit, and these little traumas can turn into increased vascularization, a little bit of inflammation locally in the rib, and this can be translated into a little bit of focal uptake on the PSMA-PET scan [inaudible 00:17:19] prostate cancer metastasis. Here the impact can be important, if you switch the patient from zero to M1 just on that, that has a major consequence for the patient management.

So in that little study that look at 60 patients with the PSMA-PET solitary rib lesion at initial staging, without any treatment, they follow these 62 patients. And you can see that only one turned out to be malignant. You can see that the uptake level was pretty low between three and two. And when you follow them after 36 months with PSA follow-up imaging or biopsy, they were all benign. So when you have only a solitary faint uptake in a rib, just take that into account, give the benefit of the doubt that it is probably nothing, and you should not preclude the patient to have a curative intent treatment locally on the prostate. Of course, if it is very intense, you have an SUV max of above, I would say, I don't know, five, seven, very focal, and you have no CT correlate that guides you towards another cause, you have to take this as a prostate cancer metastasis.

Last thing in the bone, I want to show here a little bit of when you restage patient after radiation therapy in a bone metastasis. Just that you know that the uptake of a lesion decrease slowly, you see here, after three, four months, the residual uptake in the lesion is still very high, and you need at least six months to see really a decreasing uptake. So if you restage early and you still see some uptake in the bone lesion that you know that where irradiated, it doesn't mean necessarily it is still active cancer, it just mean that you're still early in the course, and the radiation effect takes time to decrease and to see decreasing uptake on the bone metastasis. I see here a case where 12 months after you don't have uptake anymore, but at four months you still have significant uptake after 30 gray delivered by stereotactic body radiation therapy.

Let's look at now the atypical location that must be considered as pathological uptake. With such a new performing imaging modality, you relearn the patterns of spread of prostate cancer. And at the beginning, it was very hard to convince the physicians that these were really potentially prostate cancer lesions, but little by little, getting experience and getting some confirmations, we have now experience with many weird location. Lung is now pretty common, I would say, to find prostate cancer lesion in the lung. Here you even have a case of ground-glass lesion, which does not involve prostate cancer at first glance but it turned out to be a prostate cancer metastasis here. You can see here, these lymph nodes, just in front of the liver cardiophrenic area, people were not aware that prostate cancer could go there, especially in a very low volume disease patient like that. So you really have to relearn also the potential pattern of spread people are used to see with prostate cancer, because you discover a new location.

Here, for example, you have a prostate cancer metastasis in the penis, in the corpus cavernosum and spongiosum, and that were responding well under hormonal treatment as seen by the MRI here. But in very advanced patient, I've seen some brain lesion, many peritoneal lesions and all kind of things. So everything is possible.

And the last thing is the pathological uptake also be seen on in other cancers. Because again, like I said, PSMA is expressed under neovasculature cells and in many cancers, especially the one that has very high neovasculature process, you will have high uptake.

So multiple cases of other cancer entities, pancreatic, lung, rectal have been described as show here, a rectal cancer that we have. You can see the staining on the neovasculature, whereas the normal vasculature was PSMA negative. And you can see that this is clearly a focal uptake in the rectum.

And you can see here cases of hepatocellular carcinoma. It's works very well in clear-cell renal carcinoma because very vascular. So just that other cancer can also show PSMA uptake as well.

At the end, I will end up here. A little bit like always in nuclear cancer imaging, when you have something that is faint, diffuse, not really focal, maybe isolated, you have some kind of symmetric pattern when you look at the coronal images, it is uncommon location for cancer spread, you have decrease uptake if you did a late acquisition over time, or you have a CT correlate pattern that evokes another cause, this is probably not prostate cancer. On the other hand, if you have an uptake that is very intense or very focal, that you know that there is already other metastatic lesions, that when you look at the whole body level, you have some kind of asymmetry pattern, it is in a particular common location for prostate cancer spread and it increase, or it is stable over time on late acquisition on repeat imaging, then of course, it is very suspicious for prostate cancer.

Usually the temporal information is very helpful for indeterminate lesion. You just repeat the scan when you're not sure, give the benefit of the doubt, just re-monitor the PSA, do another scan in three to six months and before trying to treat right away a lesion when it is too indeterminate.

And that's something I report a lot. There is always a discrepancy between the urologists, the oncologist that needs to act on the imaging finding reported on the PET-CT reports and needs to really act if it's real, whereas the radiologist or the nuclear medicine physician, the PET-CT readers, they want to back up themselves to ensure they didn't miss a lesion. And so there is always a trade-off between specificity required by acting on the treatment lesions and the sensitivity of the PET-CT reader that must not miss a lesion. At the end, I think what matters the most is really the specificity, but I think it's always nice to have this consideration in perspectives when you deal with the indeterminate lesions.

This will be my last slide. I thank you very much. I hope you enjoy the talk and welcome to the PSMA-PET Academy program. Thank you very much.