0:02

Hello, and welcome to Noon Conference hosted by MRI

0:05

Online. Noon Conference connects the global radiology

0:08

community through free live educational webinars that

0:11

are accessible for all and is an opportunity to learn

0:13

alongside top radiologists from around the world.

0:16

We encourage you to ask questions and share

0:18

ideas to help the community learn and grow.

0:21

You can access the recording of today's

0:23

conference and previous noon conferences

0:25

by creating a free MRI online account.

0:28

You can also sign up for a free trial of

0:30

our premium membership to get access to

0:32

hundreds of case-based microlearning courses

0:34

across all key radiology subspecialties.

0:37

Today, we are honored to welcome

0:39

Dr. Steven Rowe for a lecture entitled Prostate-Specific

0:43

Membrane Antigen, PSMA, PET Interpretation and Applications.

0:48

Dr. Steven Rowe completed his medical degree and

0:50

a PhD in chemistry at the University of Michigan

0:53

before undertaking residency training in diagnostic

0:56

radiology and nuclear medicine at Johns Hopkins.

0:59

He spent seven years on the faculty at Johns Hopkins before

1:02

recently moving to the University of North Carolina, where he

1:05

continues his research in molecular imaging of GU oncology.

1:10

At the end of the lecture, please join Dr.

1:12

Rowe in a Q&A session where he will address

1:14

questions you may have on today's topic.

1:17

Please remember to use the Q&A feature

1:19

to submit your questions so we can get to

1:21

as many as we can before our time is up.

1:23

With that, we are ready to begin today's lecture.

1:25

Dr. Rowe, please take it from here.

1:28

All right.

1:28

Thank you so much for the introduction.

1:30

It's wonderful to be with you all today.

1:32

Uh, the, uh, title of my, uh, talk is, is perhaps a little

1:35

bit different than, than what may have been on the flyer.

1:38

Uh, but I think it, uh, I think it's going to

1:39

cover everything that, that we'll talk about.

1:41

So, there are certain things about

1:42

PSMA PET that we definitely know.

1:44

Uh, we have, uh, pivotal phase III clinical

1:47

trials that make us as certain as we possibly can

1:49

be in clinical medicine that PSMA PET does certain

1:52

things and works as advertised in certain ways.

1:54

Uh, there are some things that we might know,

1:57

um, various interpretive pitfalls, which we'll

1:59

talk about, uh, non-prostate cancers that

2:02

have PSMA uptake, things along those lines.

2:04

And then there are things that we need to know that

2:06

we, that we don't know yet, and that would include, um,

2:10

things along the lines of response assessment,

2:12

uh, radiomics, uh, artificial intelligence, and

2:16

we'll, we'll touch briefly on those things as well.

2:18

I, I am just recently arriving in North Carolina, so I

2:20

haven't had a chance to, to change all my slides over yet.

2:23

So, uh, you'll see it says Johns Hopkins,

2:25

and all of my slides have this kind of,

2:27

uh, backdrop of the Hopkins dome on them.

2:29

So, uh, apologies for that.

2:31

Uh, uh, the next time I, I join you for anything, uh, I

2:33

promise there'll be University of North Carolina slides.

2:38

All right, so I have a, a couple of disclosures.

2:40

I, I think the most relevant of which would be that I, uh,

2:43

I do some work with Progenics Pharmaceuticals Incorporated.

2:46

Uh, I'm a, uh, uh, consultant for them.

2:48

I'm on their, their speakers bureau.

2:50

Uh, they do, uh, manufacture and distribute, or I

2:52

suppose distribute anyway, um, uh, the PSMA agent

2:56

known as PyL, which was previously known as

2:58

DCFPyL. I do have some examples of DCFPyL,

3:02

uh, in, in this, uh, in this presentation.

3:05

Uh, but the principles all apply to, uh, to all three of

3:09

the FDA-approved, uh, radiotracers that we have for PSMA.

3:13

There's, uh, nothing specific to clarify

3:16

about, uh, uh, about the, uh, uh, the conclusions

3:20

or, uh, or any of the, uh, any of the aspects

3:23

of, of imaging that I'm going to be talking about.

3:26

All right, just real briefly, PSMA, it's

3:28

a transmembrane carboxypeptidase, it's

3:30

highly expressed on prostate cancer cells.

3:32

About 95% of primary tumors, at least, have PSMA expression.

3:36

Uh, we know that that can change over time and as

3:39

patients' tumors, uh, evolve through various treatment,

3:42

uh, modalities and, uh, and change over time, that that,

3:45

that PSMA expression may not always be maintained.

3:48

But again, at least at the, at the start

3:49

point, when patients have primary tumors, at

3:51

least 95% of those will have PSMA expression.

3:54

And there does appear to be a correlation between

3:56

expression levels of tumor aggressiveness.

3:58

That's a histopathologic finding.

3:59

I don't know that we always

4:02

perceive that correlation or that

4:05

association, uh, on the scan level, uh, because of

4:08

things like partial volume effects and perhaps other things

4:11

that are, that are at work at a macroscopic level.

4:13

Uh, but again, if you look at, uh, uh, if you look at

4:16

just expression levels versus tumor aggressiveness,

4:18

there does seem to be a relationship there.

4:20

And there are now some emerging ideas that, uh,

4:24

that simple metrics like SUVmax on a PSMA

4:27

scan may actually have prognostic significance.

4:29

And it probably comes back to this fundamental relationship.

4:33

PSMA is, as we said, is a, is a transmembrane protein

4:36

that has a large extracellular domain that has several

4:38

sites on it that we can leverage for binding of small

4:40

molecules or, uh, antibodies or even other kinds

4:44

of molecules, uh, for both imaging and for therapy.

4:48

And what we know about PSMA PET so far, and again,

4:50

these, these things, uh, everything on this slide

4:52

is at least backed up by prospective phase two data.

4:55

Uh, and most of it is backed up by prospective

4:59

phase three pivotal trials, uh, across, uh,

5:02

again, every PSMA PET agent, uh, they, they all

5:05

use sort of relatively similar trial designs.

5:08

In particular, uh, the, the two more recent

5:10

Pylarify and Pluvicto, uh, used very similar trial

5:13

designs, uh, and then gallium PSMA-11, uh, which

5:17

is sold under the trade names of Illuccix and Locametz.

5:20

Um, the, uh, the data there at least closely parallel the

5:24

data from the, the other two, uh, agents, even though the

5:26

clinical trial designs may have been slightly different.

5:30

Uh, we do know that PSMA PET has a moderate

5:32

sensitivity, only a moderate sensitivity, but a very

5:34

high specificity for preoperative nodal staging.

5:37

We know that they have a high detection.

5:38

We know that PSMA PET has a high detection deficiency for

5:40

sites of biochemical recurrence, and we know that PSMA PET

5:44

is effective for guiding metastasis-directed therapy for

5:47

patients with limited volume or oligometastatic disease.

5:49

And also for selecting patients for endoradiotherapy, uh,

5:51

which are generally, uh, lutetium-labeled ligands, at least in North America.

5:55

They're so far lutetium-labeled ligands, uh, primarily

5:56

Pluvicto, which is a regulatory-approved agent, or

5:59

lutetium-177 PSMA-617, as it's more broadly known.

6:08

All right, so here's, here's an example of

6:09

PSMA PET in a, uh, in a couple of patients

6:12

with, uh, high-risk prostate cancer.

6:14

So, and, and again to, uh, well, as we'll see over

6:17

the next few slides, this is sort of a, uh, an area

6:20

where for nodal involvement, the sensitivity is

6:22

moderate, uh, but the specificity is extremely high.

6:26

Most patients that you image with, uh, unfavorable,

6:29

intermediate, high, or very high-risk prostate cancer,

6:31

which is typically what the guideline suggests,

6:34

uh, that we should be imaging those populations.

6:37

Those patients are going to look like the patient on the left.

6:39

They'll have very high uptake in a primary

6:41

tumor in the prostate, but they won't

6:42

have any evidence of nodal involvement.

6:45

Uh, we, we know from nomograms these patients could have

6:47

potentially up to a 25% chance of, of nodal involvement,

6:50

but we also know that our sensitivity is, is moderate.

6:53

So our, uh, uh, most patients, the significant majority

6:57

of patients, even those at risk for nodal involvement, uh,

7:00

will not have evidence of nodal involvement on the scan.

7:03

Uh, my urology colleagues like to note that,

7:05

that this indicates that, uh, nodal staging

7:07

of prostate cancer patients is still ult,

7:09

still ultimately comes down to surgery.

7:11

It's ultimately surgical staging.

7:13

We can do the best we can with PSMA PET, uh, but, uh.

7:17

The surgeon is going to find disease that,

7:19

that we did not perceive on, on the PSMA scan.

7:22

Occasionally, and I would say this is uncommon,

7:24

patients will present more like the patient on the

7:25

right, where they have unsuspected systemic disease.

7:28

This patient has nodal involvement in the pelvis and the

7:31

retroperitoneum and in the left supraclavicular space, um,

7:34

this, uh, these, these nodes are all subcentimeter.

7:37

So prior to the PSMA PET scan, this patient wouldn't

7:39

have been classified as clinically localized.

7:42

Uh, but in this particular case, the

7:43

patient actually has systemic disease.

7:46

This happened a couple of times in our phase

7:47

two single-center trial, and a few times in,

7:50

uh, in the pivotal, uh, phase three trial

7:53

that was, that was done with the same agent.

7:55

Uh, but this is a relatively rare event.

7:57

However, for the patient, it's life-changing, and

7:58

that they switch from what would be a curative,

8:01

uh, attempted curative therapy, uh, to a more

8:05

palliative, uh, paradigm where they'll be receiving,

8:07

uh, presumably lifetime, uh, systemic therapy.

8:12

All right.

8:12

Let's, uh, let's quickly take a look at the,

8:15

uh, the Osprey trial, which, uh, was, was

8:17

done with the agent that sound known as ASPY.

8:20

Uh, however, very similar trial design and,

8:23

uh, very similar results with, uh, with, uh,

8:25

the other agents that are, that are out there.

8:28

Uh, this, uh, particular trial design

8:30

was sort of a follow-on to the, um.

8:33

To the trial that was, that involved both of

8:36

the patients I showed you on the previous slide,

8:38

which was again, a single-center, phase two.

8:40

Uh, the OSPREY trial was phase three.

8:42

It actually had two cohorts, but we're only going to

8:44

focus on one cohort, uh, which was, uh, a total of,

8:48

uh, um, 268, uh, uh, patients who, uh, underwent, uh,

8:54

a, uh, DCFPyL PET/CT, a PSMA-targeted PET/CT.

8:58

And of those, 252 patients underwent a radical prostatectomy

9:02

with an extended pelvic lymph node dissection.

9:05

Um.

9:06

Co-primary endpoints of the trial were sensitivity

9:08

and specificity for pelvic nodal involvement.

9:12

And the results of that trial, which again, had been

9:14

recapitulated with, with other agents, uh, in very, in very

9:18

similar clinical settings, was again, a sensitivity of, of

9:21

only 40%, which was surprising, and that primary endpoint

9:24

was missed, uh, by, by, in, in this particular case.

9:28

Uh, interestingly, the co-primary endpoint of,

9:30

uh, Pluvicto, uh, and again, a very similar clinical

9:33

trial setting that included some patients that

9:35

had, uh, uh, unfavorable intermediate risk.

9:38

These were all higher, very high-risk patients.

9:40

Uh, also met, met, also failed to meet

9:42

its, um, sensitivity co-primary endpoint.

9:47

But the unifying theme of these agents is

9:49

that the specificity remains incredibly high.

9:52

And no matter who you exclude, no matter

9:54

what sort of sub-analyses you do, post hoc.

9:57

Uh, uh, playing with the numbers, you'll generally

10:00

find that no matter what you sort of drive your

10:02

sensitivity to, the specificity remains rock solid.

10:05

This, this has important implications for how

10:08

I think we should approach reading these scans.

10:10

If you have a patient who's presenting for initial

10:13

staging, uh, and presumably those patients are generally

10:16

going to be unfavorable, intermediate, high, or very

10:19

high-risk prostate cancer, if the specificity is truly

10:23

approaches 100%, and that specificity is

10:26

robust, then in places where prostate cancer would

10:29

make sense to spread to proximal external iliacs

10:33

or, say, internal iliacs, um, presacral, perirectal.

10:38

And then if those areas are involved, we have

10:40

to start worrying about the common iliac.

10:42

And if those areas are involved, we have to

10:43

start worrying about the retroperitoneum.

10:45

But if we see disease that, that sort of fits that

10:48

paradigm of, of where, uh, prostate cancer spreads to.

10:53

Our specificity for that uptake is going

10:55

to be incredibly high, and we should

10:56

confidently call even very subtle findings.

11:00

Uh, that's my approach.

11:01

If I see anything above blood pool in those areas

11:03

that we just described, um, and I can correlate

11:06

that to a lymph node, even if it's very small, I'm

11:09

highly suspicious that that lymph node is involved.

11:11

I'd say the, the caveat to this is you will

11:13

sometimes encounter patients who have systemic

11:16

inflammatory processes, and they'll have relatively

11:18

diffuse adenopathy with low-level uptake.

11:21

And I think those patients are an important pitfall to

11:24

at least be aware that, uh, uh, that someone with diffuse

11:28

adenopathy with uptake, uh, may be fooling you and may be

11:31

in that couple of percent of patients where, where the

11:33

specificity, uh, doesn't quite hit 100%.

11:38

All right, so we also said that we know that, uh,

11:40

the PSMA PET has a high detection efficiency

11:42

for, for finding sites of biochemical recurrence.

11:45

Uh, here you can see a local recurrence

11:46

that was occult on, uh, occult on CT.

11:49

Uh, but there's an enhancing nodule visible there on MRI.

11:53

Even more commonly than that, patients will have one

11:56

or perhaps two, uh, lymph nodes in the pelvis that,

11:59

that are, that indicate their signs of recurrence.

12:01

Uh, the pitfall here is trying to avoid, uh, mistaking

12:04

ureteral excretion for uptake in the lymph node.

12:08

That can be challenging, and it's really incumbent upon us

12:10

to look very carefully and make sure that we're not over-

12:13

calling, uh, ureteral excretion as a, as a pelvic lymph node.

12:18

The, uh, CONDOR trial was, was now one of

12:20

several clinical trials that had been done

12:22

in this space with various PSMA agents.

12:24

Uh, SPOTLIGHT was the, uh, uh, was the PSMA

12:27

equivalent, and then there was also a, uh,

12:29

a dual-center study between UCSF and UCLA.

12:33

Uh, looking at this, uh, same patient population or a

12:35

similar patient population, uh, with gallium PSMA-11.

12:42

The results of the CONDOR trial, I think also sort of

12:44

informed how, how we should approach the scan, or not

12:46

just the CONDOR trial, but, but all of these trials

12:48

have, uh, have generally found similar, similar results.

12:52

The big result is, uh, is actually on the

12:55

next slide, but I do want to, uh, I do

12:56

want to focus on the slide for just a second.

12:58

One thing the FDA has done is PSMA agents have

13:01

been approved is start to, uh, is start to have all

13:05

these new metrics, uh, correct localization rate.

13:07

Correct detection rate.

13:09

These generally come down to being some variation

13:11

on a theme of positive predictive value, and usually

13:15

at the lesion level, although sometimes not quite at

13:17

the lesion level, sometimes more of a patient level.

13:20

Uh, and it's, uh, honestly, the, the

13:22

FDA hasn't done us any favors here.

13:24

They've, they've just kind of muddied

13:25

the waters and made things confusing.

13:27

However, uh, for, uh, uh, for

13:30

this particular trial, the, um.

13:33

The primary endpoint was something called the

13:35

correct localization rate, which was a, uh, a

13:39

positive predictive value at the lesion level,

13:41

is at least how we should think about it.

13:43

And we won't talk about the standard of truth, but

13:45

there was sort of a hierarchical standard of truth.

13:47

The correct localization rate does

13:49

not change much with PSA level.

13:51

So that's, uh, it changes a little bit.

13:53

There's a trend towards higher correct localization

13:55

rate at higher PSAs, but it's not nearly

13:59

as strong as the, uh, as the relationship

14:02

between, uh, detection efficiency and PSA level.

14:06

Again, what I think this means for how we should

14:08

approach PSA scans in a biochemical recurrence

14:10

setting is that we know that at very low PSA levels

14:13

that many of those scans are going to be negative.

14:16

However, if our correct localization rate, or positive

14:19

predictive value at the lesion level, is relatively

14:22

robust across PSA levels, the same finding may mean

14:26

the same thing regardless of the patient's PSA level.

14:28

And what I mean by that is a small re fossil

14:31

lymph node in a, in a biochemical recurrence

14:33

patient with a PSA of 0.3 or a PSA of three.

14:38

And we treat that finding in the same way and be highly

14:40

suspicious that it represents a true positive site

14:42

of disease because that positive predictive value at

14:46

the lesion level is, is so high regardless of PSA.

14:49

So that's at least my approach to,

14:51

uh, to the, uh, the scans here.

14:53

Uh, but again, I maybe the most important take-

14:55

home message is that PSMA isn't a magic bullet.

14:57

And at low PSA levels in the recurrent setting,

15:00

it is not uncommon for a scan to be normal.

15:04

Now, if we, if we do assume that, uh, that these, uh, that.

15:09

We have a high detection efficiency at, uh, for most

15:12

biochemical recurrence patients, and that we can see

15:14

things that we can't see with conventional imaging.

15:17

The next step down that line is to say, well,

15:19

if we see a limited volume of disease, is that

15:20

patient eligible for metastasis-directed therapy?

15:23

Uh, perhaps avoiding, or at least

15:25

kicking the can down the road on, uh.

15:28

On systemic therapy like androgen deprivation, which, uh,

15:32

in addition to things like, uh, quality of life and sexual

15:35

side effects, also puts men at risk of, um, heart attacks

15:39

and strokes and other, other sort of bad complications.

15:41

So, uh, so we'd like to avoid that if we can.

15:44

Of course, most men aren't going to be able to

15:46

avoid that, but it occasionally we can at least, uh,

15:49

perhaps buy a couple of years off of systemic therapy.

15:52

Here is a patient who, after prostatectomy, had what

15:54

appears to be a, a presacral or perirectal lymph node, that

15:57

was the only visible site of disease on a PSMA scan.

16:00

Uh, the patient, uh, underwent, uh, stereotactic

16:02

body radiation therapy to that site of

16:04

disease, did not start systemic therapy.

16:07

His PSA became undetectable.

16:09

And over the two years of follow-up of, of this particular

16:11

study, uh, his PSA never became detectable again.

16:15

I have absolutely no idea if this patient's cured.

16:17

I, I would be hesitant to use cure in anyone

16:19

who has disease outside of the prostate, but

16:21

at least this patient's been at least two years

16:23

without requiring androgen deprivation therapy.

16:27

We also see things at the other end of the

16:28

spectrum where, again, the lack of sensitivity

16:31

of conventional imaging may fool us.

16:32

Uh, here was a bone scan that had

16:34

maybe one suspicious spot on it.

16:36

Uh, but, uh, in fact this patient had

16:38

hundreds of marrow-based lesions that were,

16:41

uh, invisible on both CT and bone scan.

16:44

So this patient obviously not eligible

16:46

for metastasis-directed therapy.

16:47

Uh, this patient's volume of disease requires

16:49

that, uh, that he undergoes systemic therapy.

16:52

And there are patients that sort of fall in the middle.

16:54

They have low-volume disease, but

16:56

perhaps not as low volume as we'd like.

16:58

Uh, this patient, for instance, has both pelvic and

17:00

retroperitoneal adenopathy and underwent SBRT without

17:04

starting, uh, without starting ADT, uh, and returned

17:07

a few months later for a follow-up PSMA PET scan and,

17:09

and was found to have disseminated bone metastases.

17:12

So those bone metastases were almost certainly there

17:15

prior to him undergoing radiation therapy, but they

17:17

were perhaps below the detection efficiency of the scan.

17:19

So again, PSMA not a magic bullet.

17:21

Uh, it will serve some patients well in

17:23

selecting them for metastasis-directed therapy.

17:25

Uh, in other patients, uh, it may not, uh, it may not start

17:28

so well because they may just have such low-volume disease,

17:31

if you will, the true micrometastatic disease, uh, that

17:33

we're just not able to treat all of their sites of disease.

17:38

However, if a, if a man has sort of progressed to

17:41

the point that he has more widespread metastatic

17:43

disease, and in particular in the US this would mean

17:45

post-chemotherapy, metastatic castration-resistant

17:48

disease, uh, to be eligible for treatment with lutetium,

17:51

PSMA, or again, in the US right now, that's Pluvicto.

17:55

Uh, these, uh, uh, we, we select patients, uh, for

17:59

Pluvicto with, uh, uh, with a PSMA scan upfront,

18:03

uh, so that we can confirm that they have both

18:05

sites of disease that, that have, uh, that are

18:07

expressing the target and also that they don't have

18:10

sites of disease that aren't expressing the target.

18:11

And that won't benefit from, from

18:13

PSMA-targeted endoradiotherapy.

18:16

Uh, somewhere around 40% of patients that are treated

18:19

are going to have an objective biochemical response

18:21

with a, uh, greater than 50% drop in their PSA.

18:25

There would be about 70% of patients

18:27

that have any drop in their PSA.

18:29

Uh, and that efficacy all comes in, in a, uh, in

18:33

a relatively small cost in terms of toxicities.

18:37

Uh, there can be, uh, nephropathy, but

18:39

it's generally low grade and self-limited.

18:42

There can be xerostomia, again,

18:44

generally low grade and self-limited.

18:47

And we now have sort of the pivotal

18:48

trial results with Lu-PSMA-617.

18:51

Uh, this was the VISION trial published in

18:53

the New England Journal a couple of years ago.

18:55

Uh, and it showed that in a prospective multicenter

18:58

setting that, uh, there was, uh, improvement in

19:01

progression-free survival, freedom from skeletal events,

19:04

and then the big one, of course, overall survival in

19:06

those patients who, uh, received, uh, Pluvicto on top

19:10

of best standard-of-care therapy relative to those

19:13

men who only received best standard-of-care therapy.

19:18

Let me, uh, let me shift gears a little bit from

19:20

what we know to what we think we might know.

19:22

And, uh, I think there's, uh, there's a drive right now.

19:26

Uh, but this is one of my, my academic interests and,

19:29

uh, but it's also academic interests of other folks

19:31

that, um, we, we need to, we need to do a better job

19:37

of, of sort of bringing in, in a, into alignment,

19:40

uh, everyone's interpretations of the same findings.

19:43

Uh, you'll find a lot of inter-reader, uh, a

19:46

lot of interrater variability right now in terms

19:48

of how people approach findings on, on scans.

19:50

A solitary rib metastasis, a solitary rib lesion

19:52

may be a metastasis to one reader, and it

19:54

may be fibrous dysplasia to another reader.

19:57

And of course, all of this is going to eventually,

19:59

uh, hopefully shake out as, as we collectively

20:02

become more experienced and all read hundreds or

20:04

even thousands of these scans over the years, but

20:08

we, uh, uh, we do need some way of

20:10

homogenizing our interpretations.

20:12

We need to communicate findings to our clinical

20:13

colleagues, uh, and other fields have already done this.

20:17

And, and PI-RADS is probably the big one.

20:19

Uh, I, I sort of came into being as a radiologist

20:22

when PI-RADS already existed, but all the stories

20:24

I hear about, uh, sort of the pre–PI-RADS era

20:27

was that, uh, uh, you know, no one, uh, uh,

20:30

everyone was reading prostate MRI differently.

20:32

Everyone was interpreting what those reads

20:34

meant differently, and it was hard to sort of

20:36

get a nice, coherent signal out of that noise.

20:39

And PI-RADS sort of brought everything into alignment,

20:41

where now the expectation is everyone's reading.

20:44

Uh, with PI-RADS, there may still be some inter-

20:46

reader variability, of course, but at least PI-RADS

20:48

has provided sort of a structure that, that, uh, that

20:51

can, uh, uh, on which everyone can hang their hat.

20:54

Now there are a number of.

20:55

Competing approaches, but I'm only going to talk about why

20:58

that's PSMA-RADS, uh, since it's kind of like BI-RADS or

21:01

at least, uh, uh, is, is potentially like BI-RADS or PI-RADS.

21:05

And then it's a, a Likert scale of, of, uh, how,

21:08

uh, likely prostate cancer is in given patients.

21:12

But let me first show you some pitfalls as to

21:14

why this would be an important thing to do.

21:17

Uh, obviously if everything, uh, is that's hot on the

21:19

scan is cancer and everything that isn't hot on the

21:22

scan isn't cancer, uh, our lives would be a lot easier.

21:25

Uh, but of course there are pitfalls in both directions.

21:28

For example, a, uh, a scan that may have

21:30

negative or variable uptake, uh, as in this

21:33

patient, uh, may indicate that a patient has

21:35

neuroendocrine prostate cancer. Neuroendocrine

21:37

prostate cancer tends to lose PSMA expression.

21:41

And so you can get what are, uh, absolutely biopsy-

21:44

proven positive liver metastases in a patient

21:47

that doesn't wind up having any PSA uptake.

21:51

Uh, and this is just the, the biopsy evidence of that.

21:53

So, uh, adenocarcinoma, uh, has PSMA expression, but

21:58

as you can see in those central couple of panels,

22:00

all that PSMA expression goes away in the sort

22:02

of truly neuroendocrine-differentiated prostate

22:05

cancer that was present in the liver, this patient.

22:09

Uh, there are other things that are probably neo-

22:12

vascularized or inflamed that can have PSMA uptake.

22:16

Uh, we don't necessarily always understand

22:17

the biology of this, but here's one example.

22:19

This is a patient with a vertebral compression fracture.

22:22

This, uh, uh, this has nice linear uptake.

22:25

It looks a lot like it would on, say, an FDG PET

22:27

scan, uh, but no fatality here to suggest that this

22:30

is a, uh, a, a pathologic compression fracture.

22:36

Maybe one of the more famous pitfalls in, uh, in PSA

22:39

imaging is the presence of, uh, peripheral ganglia.

22:44

So, dorsal root ganglia, cervicothoracic, or stellate ganglia,

22:48

and celiac ganglia are all variably positive in patients.

22:53

And the celiac ganglia is, is really the hardest one.

22:55

So that's the panel in the upper right there.

22:57

A couple of panels in the upper right, it

22:59

looks like a retroperitoneal lymph node.

23:01

It looks, for all the world, like a retroperitoneal

23:03

lymph node, and in some patients it's quite prominent.

23:05

In some patients it's quite avid.

23:07

So it really, it's incumbent upon us to know the

23:08

anatomy well and not mistake these lesions for,

23:12

uh, or not mistake these normal structures with

23:14

variable uptake for sites of prostate cancer.

23:18

Now here's just another example of that.

23:21

And here, here's that, that celiac ganglion.

23:23

So again, it really does look like a lymph node.

23:25

It's kind of kidney-bean shaped.

23:26

Uh, and this patient's relatively prominent,

23:28

and they have relatively prominent uptake.

23:30

Uh, but again, we just have to be very careful not to,

23:32

uh, not to mistakenly call those sites of prostate cancer.

23:36

Uh, and then here's, uh, here's an example of the bone.

23:38

The bones are, are full of hazards for

23:40

those of us interpreting these scans.

23:42

Uh, here's a patient who had Paget disease in the sacrum,

23:46

which, uh, which had uptake throughout the sacrum.

23:48

Uh, you can see the bone scan there

23:50

also had uptake throughout the sacrum.

23:52

Uh, but, uh, uh, but the MRI shows, uh, shows that

23:55

there's preservation of the intramedullary fat.

23:58

So this is just a nice example of Paget disease.

24:02

And as noted, inflammatory things can, can have uptake.

24:05

So here's a, a patient with cortical laminar necrosis

24:07

after getting radiation therapy to the brain.

24:09

Uh, you can see that there's uptake in the, uh,

24:12

areas of enhancement, uh, on the, uh, on the MRI.

24:16

So this, uh, uh, this uptake is,

24:19

uh, is potentially confusing to us.

24:21

You'll see this in other parts of the body as well.

24:23

Patients who have gotten prior SBRT, uh, to, say, a rib

24:26

lesion, the underlying lung, the post-radiation changes

24:29

there, uh, will often have low-level diffuse uptake.

24:33

And then there's PSMA expression in non–prostate cancer.

24:35

So this is primarily in the tumor neovascularization, the neovascular endothelium.

24:37

Uh, but, uh, while we can potentially leverage this in

24:40

these other cancers for diagnostic purposes, we also have

24:43

to really be aware that if a pattern of apparent disease

24:46

does not, uh, comport with what we believe prostate

24:49

cancer should look like, we may have to suggest that

24:52

another prostate or then another cancer type is present.

24:55

Let me show you a couple of examples of what used

24:58

to be called conventional renal cell carcinoma.

25:00

We now call clear cell renal cell carcinoma.

25:02

563 00:25:05,580 --> 00:25:07,110 Uh, and here's a patient with

25:07

relatively widespread metastases.

25:08

Perhaps the clue on his whole-body

25:10

map images that he's missing a kidney.

25:12

Uh, but these were all, uh, these were all

25:14

metastatic renal cell carcinoma, uh, and

25:17

specifically clear cell renal cell carcinoma.

25:19

And you can see even very, very small lesions there.

25:21

There's quite avid uptake.

25:23

But I would, uh, I would posit that something

25:25

like a, uh, subcutaneous perineal metastasis

25:29

would be highly unusual in prostate cancer.

25:32

Something like an intramuscular metastasis

25:34

would be highly unusual in prostate cancer.

25:36

But those things may indicate the presence of,

25:38

of something like a renal cell carcinoma that

25:40

has more unusual patterns of disease spread.

25:42

Uh, brain parenchymal metastases would

25:44

also be very unusual in prostate cancer.

25:46

Uh, here is one from, uh, from, again,

25:48

a clear cell renal cell carcinoma.

25:52

Then, relative to FDG, I tend to think

25:54

FDG gets kind of a bad rap for RCC.

25:56

It's not bad in the metastatic setting, uh, but in a patient,

25:59

uh, contemporaneously imaged with, uh, with FDG and PSMA,

26:03

uh, the PSMA not only had higher uptake in known lesions,

26:06

but also, uh, uh, but also, uh, wound up, uh, picking up a

26:10

couple of lesions that would have been missed on the FDG scan.

26:14

And then one last example of a clear cell RCC.

26:16

This is a patient who was imaged, uh, right before he passed

26:20

away, unfortunately, and then underwent a rapid autopsy.

26:23

And so, uh, sites of disease that would have potentially been

26:26

very hard to biopsy, such as intramuscular metastases, uh,

26:29

were available for, for histopathologic inspection.

26:33

So not only does, uh, PSMA do a sensitive job of

26:37

picking up clear cell renal cell carcinoma metastases, uh, it

26:39

also is very specific for, for those same metastases.

26:43

Uh, I would say it is, uh, uh, again, impor, uh,

26:46

perhaps most important for those of us reading

26:48

this, this, uh, modality for prostate cancer.

26:51

Just to be aware that very strange patterns of spread

26:53

from prostate, what look like strange patterns of spread,

26:56

spread from prostate cancer may be something else.

26:58

And again, we, we have to be aware,

27:00

we do have to be aware of that.

27:02

Uh, let me show you, uh, a couple of, uh,

27:05

these are sort of high-grade glioma examples.

27:07

I, I don't think this is a specific finding.

27:09

As you saw, post-radiation changes can also have uptake.

27:12

So, uh, uh, so while not specific,

27:15

uh, this, uh, areas of blood-brain barrier

27:18

breakdown will be highlighted by PSA scans.

27:23

There's at least some uptake of

27:26

PSMA in, uh, in thyroid cancer.

27:29

Uh, I've overall been, however, disappointed with

27:31

our experience with trying to image these patients.

27:34

Uh, there seems to be both disease that's

27:36

iodine-avid, PSMA-negative, uh, disease

27:40

that is iodine-avid, PSMA-avid, and disease

27:43

that is, uh, I, uh, uh, let's see.

27:46

I forget even what I said.

27:47

Uh, that's, uh, iodine-negative, PSMA-avid.

27:50

So all those things, all those permutations can exist.

27:53

So I don't know that there's a great role for,

27:55

for PSMA PET moving forward in thyroid cancer.

27:58

But all of this brings us to the idea that, again,

28:00

we have to have ways of either, uh, structuring our

28:04

interpretation of what we think these various things

28:06

are, these various pitfalls or non-prostate cancers.

28:10

And we have to communicate that to our clinical

28:13

colleagues, uh, and help them sort of make decisions

28:16

as to, uh, what new workup, uh, what new workup steps

28:20

might be for the patient if it's not so straightforward.

28:22

It's just they have prostate cancer

28:24

in this site, in that site.

28:26

Uh, in our clinical, uh, in our clinical, uh, workflow,

28:29

uh, we use PSMA-RADS, uh, I suppose to some degree.

28:32

This is our bias.

28:33

We, we, uh.

28:35

Sort of developed and validated this system.

28:37

So, so it does tend to be something that,

28:39

uh, that we're very comfortable with.

28:41

Um, there, there are other systems out there, uh, PROMISE.

28:45

Uh, there's an original sort of EANM

28:47

Delphi consensus model.

28:50

And then there's something called E-PSMA, which

28:52

tried to subsume all of those things together.

28:55

Uh, but, uh, wound up, uh, perhaps

28:57

becoming a little overcomplicated.

28:59

And then, uh, there are new efforts.

29:01

There's now a PSMA-RADS 2.0.

29:02

There's also a PROMISE 2.0, uh, and all, but

29:06

all these things have the same goal of trying to

29:08

standardize some aspect of our approach to the scan.

29:10

Uh, I'll very, very briefly show you sort of, uh, what

29:14

we, uh, what we've done with, uh, with sort of PSMA-

29:17

RADS version one, uh, which was really just a, uh.

29:21

A way to, again, create a five-point

29:24

scale of likelihood of prostate cancer.

29:26

One is definitively not prostate cancer.

29:29

Uh, either from there not being any findings or

29:32

from things that do have uptake, uh, being either

29:35

biopsy-proven or pathognomonic, shown on some other

29:38

imaging to, uh, not be sites of prostate cancer.

29:42

PSMA-RADS 2, things that are very unlikely to be prostate

29:44

cancer, uh, but perhaps a low-level uptake in something

29:48

that could be obscuring a site of prostate cancer.

29:52

PSMA-RADS 3 has subcategories.

29:55

Uh, and those include, uh, RADS 3A for, uh,

29:58

indeterminate soft tissue.

30:01

RADS 3B for indeterminate bone.

30:03

Uh, just because of the workup, uh, the sort of workup

30:06

steps for each of those might be a little different.

30:07

That's why we subcategorize them.

30:10

And then things that, uh, uh, PSMA-RADS 3C are things that

30:13

might be other cancers, uh, but that have uptake.

30:17

Uh, here you can see an example of a potential lung cancer.

30:19

Uh, this was an isolated finding in this patient.

30:22

We will always, within reason, try to get

30:24

tissue on isolated findings in the lungs.

30:27

Uh, although they often do turn out to be prostate

30:29

cancer, that is still a relatively rare pattern

30:31

of disease spread for prostate cancer.

30:33

So our preference is to make sure

30:35

that it isn't an underlying lung cancer.

30:37

Uh, you can see there are also things that might

30:39

be cancer, such as a relatively large pulmonary

30:41

nodule that lacks uptake of PSMA-targeted agents,

30:45

and those we would categorize as a PSMA-RADS 3D.

30:49

And then the fours and fives are easy.

30:50

Fours are things that are, uh, almost certainly prostate

30:53

cancer, but don't necessarily have a CT correlate.

30:56

Might be a subcentimeter lymph node; it might be

30:58

a bone lesion that's marrow-based and doesn't

31:00

have an associated sclerotic reaction.

31:02

And then those things that, uh, are either things

31:05

like enlarged lymph nodes or sclerotic bone lesions

31:08

that have uptake, uh, those go into the five category.

31:12

So it's, it's a relatively simple approach, uh, but

31:14

it does make you come down hard on sort of giving some

31:17

kind of number or decision to things that, uh, that

31:20

may, uh, uh, where it may have to make a decision.

31:22

Does this need additional workup,

31:24

or is this definitively benign?

31:25

Or is it, uh, is it not definitively benign?

31:29

All right, this is just an example.

31:31

Uh, I won't, won't go into the details here.

31:33

Um, this, uh, this approach has a, uh, has a higher

31:36

interobserver reliability, uh, both on sort

31:39

of an overall scan score, highest score for, for an

31:43

individual lesion, as well as individual lesion scores.

31:46

These aren't perfect, uh, and experience does matter.

31:50

Experienced readers will tend to cluster together a little

31:53

bit more, but relative to something like PI-RADS, the,

31:56

the repeatability here is actually substantially higher.

31:59

And those things that are indeterminate,

32:01

unfortunately, really are indeterminate.

32:02

So if it's an indeterminate lymph node, uh, when you follow

32:06

it up over time, maybe about 75% of those will manifest

32:09

as, as having been true-positive, and maybe 25% won't.

32:13

Uh, and, uh, bone lesions, it's actually worse.

32:16

So with bone lesions, the, uh, uh, only about 20%

32:19

of indeterminate bone lesions wind up, uh, wind

32:21

up being, uh, uh, being true-positive on follow-up.

32:26

And advanced reconstruction

32:27

algorithms can help a little bit.

32:28

Here's just a patient who is imaged with and without,

32:31

uh, a, uh, point spread function reconstruction.

32:33

And you can see a, a lymph node that may be a

32:36

little wishy-washy, although probably callable, uh,

32:39

in the top couple of panels, uh, is, uh, is more

32:42

definitively callable in the, uh, in the bottom panels.

32:46

All right, so let me, let me move on to some emerging ideas.

32:49

So these are things we, we don't know,

32:50

but, but hopefully will at some point.

32:53

And there are, uh, there are things here

32:55

I'll probably go by pretty quickly, uh, make

32:57

sure we, we have time for, for questions.

32:59

'Cause we've probably covered sort of the, the really

33:01

practical aspects of things up, up to this point.

33:03

And now we're, we're going to, uh, do some thought exercises

33:06

about maybe where the field of, of PSMA PET is headed.

33:08

So let me, uh, but let me launch, uh, right into things.

33:12

Uh, so very briefly, image quantitation.

33:15

So I'm a big believer in image quantitation

33:16

from the perspective of biomarker development.

33:19

I don't think there are meaningful PSMA SUV

33:23

cutoffs that we can really apply.

33:26

But I can tell you that, uh, liver uptake is highly

33:30

reproducible, uh, across various radiotracers.

33:33

Uh, they'll all have different liver uptake, but the liver

0:33

uptake is sort of the most repeatable organ in, in PSMA PET.

33:39

Uh, there's also, there's an agent over in

33:41

Europe called PSMA-1007 that has very high

33:44

liver uptake because it's hepatobiliary excreted.

33:47

Uh, so that, I actually don't know if the,

33:49

if the quantitation is most repeatable in the

33:51

liver, but for the agents that are FDA-approved

33:53

in the US, this, uh, this would hold true.

33:56

Uh, there.

33:57

Probably is some sort of tumor sink effect.

33:59

When you see really extreme degrees of, of tumor

34:03

infiltration, particularly in the skeleton, there's

34:06

definitely a drop in uptake in other organs, but

34:09

it doesn't really come into play at sort of the,

34:12

uh, the levels of disease that we often see here.

34:15

You can see three patients who have varying levels of,

34:18

of disease involvement, but it really looks like their,

34:20

their organs are all very similar in degrees of uptake.

34:24

Again, there's a little bit of a tumor sink effect, but

34:25

it is, uh, there are other things that affect PSMA uptake

34:29

that that would also play a role in sort of the day-to-day

34:32

variability, so it won't be all tumor effect, uh, or just

34:36

differences in sort of tumor volume between patients.

34:40

Uh, let's see.

34:41

I think I'm going to skip that.

34:43

Uh, but here, uh, uh, but here I'm going to, I'm

34:46

going to maybe slow down just a little bit because I,

34:48

I think here we have a couple of important ideas.

34:51

So, uh, we know that, uh, we know that,

34:54

uh, test-retest studies are not easy to do.

34:57

Uh, patients, uh, getting imaged.

35:00

You know, at one time, and then again, you know, within

35:02

a week without any intercurrent therapy or, or initiation

35:06

of new therapy, uh, it's hard to sort of line all that up.

35:08

It's expensive to do that.

35:10

It's lots of scanner time.

35:11

But, uh, we, uh, we're lucky enough to have the

35:13

Prostate Cancer Foundation, uh, fund us to do a test-

35:16

retest trial, uh, with, uh, patients with metastatic

35:20

disease, uh, with, uh, with a PSMA PET scan.

35:23

And it turns out that, uh, that

35:26

the scans are highly repeatable.

35:28

So, uh, even in low-volume disease or in high-

35:30

volume disease, uh, the SUVs are, are robust.

35:34

Uh, and, uh, here you can see some

35:36

Bland–Altman plots and correlations.

35:38

Uh, the—

35:40

I will say that the, the SUVs

35:42

are more robust at higher SUVs.

35:45

So as you get these patients who have metastatic

35:47

castration resistance, extensive volumes of disease,

35:50

and very high SUVs, those SUVs will be very repeatable.

35:54

So the idea that, uh, a, a solitary PSMA scan at,

35:58

at, at a sort of a time point prior to initiation

36:01

of something like Pluvicto, uh, is going to tell you

36:03

something about the dosimetry of those patients

36:06

and how much dose their tumor is going to get.

36:09

I think there's a lot of validity to that idea.

36:11

I don't know that we're leveraging that very

36:13

well, but hopefully we will at some point.

36:15

But, uh, but again, yeah, higher

36:16

SUVs are more robustly repeatable.

36:19

Uh, and I think that's in and of

36:20

itself an interesting finding.

36:22

The SUVs in FDG PET just don't have the

36:24

dynamic range to actually have told us that,

36:27

although it's relatively intuitive, perhaps.

36:29

Uh, but we, we didn't know that prior to the advent

36:31

of, of PSMA PET, uh, where we would have SUVs

36:34

from down at, uh, you know, 0.8 or 0.9 sometimes

36:38

in early subtle sites of disease, up to over a

36:40

hundred in, uh, in really avid sites of disease.

36:45

All right, this is just more about that.

36:47

Okay.

36:48

So radiomics has, has gotten a lot of play, and

36:50

it would be great if radiomics could perhaps, uh,

36:52

provide us some insights or predictive biomarkers,

36:55

uh, that might tell us something about PSMA PET.

36:58

Unfortunately, I don't tend to be a huge

37:01

believer in radiomics for PET in general.

37:04

Uh, there are broad categories of things that tend to

37:07

be, uh, sort of low-frequency features of the images.

37:10

Things like entropy or homogeneity, uh, that

37:13

do have some repeatability or reproducibility,

37:16

uh, at the, uh, at the radiomics level.

37:19

Uh, I'll also say, though, that, uh, um, that most radiomic

37:23

features are not reproducible, and that has to do with,

37:26

with aspects of the, the PET reconstruction matrix, where

37:29

there are, uh, inestimable parameters that are high-

37:34

frequency features of the images that are basically going to

37:37

never be repeatable because we're just throwing random

37:39

numbers in that, uh, in that reconstruction matrix.

37:43

I, uh, probably need to update this,

37:45

if you would like to read more.

37:46

Uh, this, uh, manuscript has now been published

37:48

in, uh, uh, in, uh, The Prostate, uh, Rüdiger Veer,

37:53

who's done a lot of, uh, a lot of this work.

37:54

Uh, uh, who's over at Würzburg, uh,

37:57

is the first author on that paper.

37:58

It, it's worth a read; it's kind of an interesting

38:00

fundamental PET, uh, fundamental PET finding that

38:04

that just happens to be applied in this case to PSMA.

38:08

And artificial intelligence.

38:09

You know, I think every radiologist worries that at

38:11

some point, uh, these algorithms are gonna get so good

38:14

at looking at images that we're gonna be outta jobs.

38:16

Uh, I, I think our, our jobs are hard enough that

38:19

that isn't, like, right on the horizon, but, uh,

38:22

but it is certainly going to come to pass that AI

38:24

is gonna replace a lot of what we do day to day.

38:26

So we do have to at least have that

38:28

in the back of our minds right now.

38:30

AI, I think, is gonna be more of a friend to us.

38:32

It's gonna help us with lesion classification.

38:34

We can do whole-body tumor burden assessments, uh, and

38:37

prognostication and decision making will, will hopefully

38:40

be things that we can start to lean on AI for a little bit.

38:45

Uh, with PSMA PET, AI is already pretty good.

38:47

I, I think it's better at hotspot detection

38:50

than it really is at classification.

38:52

Uh.

38:53

The, the FDA-approved product that's out there, that's,

38:56

that's a sort of PSMA AI tool, uh, will pick up solitary

39:00

rib lesions that are fibrous dysplasia and things like that.

39:03

It's basically picking up things that are

39:04

outside of the normal biodistribution.

39:06

It does a very good job at that.

39:07

It provides a quantitative readout, uh, but it isn't

39:10

at the point where it's helping us make decisions about

39:13

what the lesions really are or what the sites of uptake

39:16

really are and how seriously we need to take them.

39:19

That's really where we have to go for AI to ultimately

39:22

be useful for us in this, this context.

39:25

Uh, but, uh, but it, uh, isn't quite there yet.

39:28

I think it's, it's a question of data and

39:30

follow-up and, uh, and really good data.

39:34

So data that is based on, uh, uh, that's based on either

39:38

long-term follow-up or, or histopathology correlations, and

39:42

that kind of data is sort of hit and miss in the field.

39:47

I, uh, I always wonder if we can

39:48

supercharge some of these sort of, uh, um.

39:51

These very image-oriented

39:54

algorithms with even better images.

39:58

Uh, this is just an example of cinematic rendering

40:00

of, of an PSMA PET/CT, which is, uh, uh, which

40:04

is a, a technique that was pioneered by Siemens.

40:07

Uh, it's available on their VB40 and above workstations,

40:11

I think.

40:12

Uh, and, uh, this part isn't necessarily

40:15

out there and ready for prime time yet.

40:17

Uh, to differentiate the PET data from the CT

40:19

data, you have to internally light the PET data.

40:21

Uh, but nonetheless, I'd, I'd really be curious

40:23

if, uh, if, uh, GPUs can sort of derive more out

40:28

of these, uh, interesting and color-coded, uh,

40:33

and sort of, uh, 1:1 scan and one-image kind of

40:38

image, uh, that they can, out of sort of just the

40:41

2D images being, being generally fed to them.

40:45

We, uh.

40:47

In addition to, uh, in addition to thinking about

40:49

radiomics and AI, uh, we need to also think just

40:52

more broadly about imaging biomarker development.

40:55

I, I think this is, this is where a

40:57

lot of the field of PSMA is headed.

40:59

Uh, again, some of that'll be driven by, by AI

41:01

and, uh, and perhaps by radiomics to some degree.

41:04

But, but we could do simpler things than that.

41:08

We can, we can look at, uh, say, in this

41:11

particular, it was a very nice study from,

41:13

uh, from the Netherlands and Australia.

41:15

They looked at patients who are PSMA-positive,

41:19

uh, on the scan, uh, histopathology-positive,

41:23

or I'm sorry, histopathology-positive.

41:25

And then patients who are negative on

41:27

the scan and histopathology-positive.

41:29

So essentially false-negative scans.

41:31

Those false-negative scans, those patients did a lot better.

41:34

So the false-negative scan is a prognostic biomarker of

41:38

how the patient's going to do with surgery and how long

41:40

their biochemical recurrence–free survival will be.

41:45

Uh, another example perhaps of an imaging

41:46

biomarker is that if we, uh, if we look at the

41:49

phase two ORIOLE trial, this is a post hoc analysis,

41:52

but nonetheless, a, a nice post hoc analysis.

41:54

It shows that if you, uh, if you treat everything that's

41:57

PSMA-avid, uh, and visible on conventional imaging versus

42:02

everything that's only visible on conventional imaging,

42:05

the patients who got everything that was PSMA-avid treated

42:07

did better than those patients who may have had PSMA-

42:10

avid disease that wasn't visible on conventional imaging.

42:14

Not a surprise, perhaps, but shows us that, uh, that we have

42:18

to be using sensitive imaging if we're going to do this.

42:20

And then if we use that sensitive imaging, we know

42:22

that those patients are generally going to do very well.

42:27

Uh.

42:29

Now, uh, maybe switching gears a little bit, but, but

42:31

keeping in the sort of imaging biomarker space, uh,

42:36

initial experience with, with PSMA in response to ADT.

42:40

So there's a, uh, poorly understood biology of

42:44

increasing PSA expression, uh, with

42:47

a decrease in androgen signaling.

42:50

That's at least what happens in the lab, is whether

42:52

that really happens in patients is hard to say.

42:55

Uh, here is a patient in whom that did seem to

42:56

hold true, where the initiation of ADT, uh, brought

42:59

out new lesions and made other lesions more avid.

43:02

It seemed like this is going to be a way that we could,

43:05

um, all of a sudden now give Pluvicto after a shot of

43:08

ADT, uh, and the patient's going to respond really well.

43:11

Or if we're trying to do metastasis-directed

43:13

therapy, maybe we find one new site of disease and,

43:16

and that changes, uh, how we treat the patient.

43:18

But, uh, but it turns out it's

43:20

actually much uglier than this.

43:21

Uh, and so if, if you look at sort of a

43:23

different context of cutting off androgen

43:25

signaling with abiraterone or enzalutamide, uh.

43:29

Patient scans do all sorts of things.

43:30

Some get hotter, some get colder,

43:32

some have new disease, some don't.

43:34

Uh, but you can sort out on kind of a whole-

43:36

body level what the, what the imaging biomarkers

43:38

are, and they have associations with, uh,

43:41

time to therapy change and overall survival.

43:44

So it's exciting to think that our changes in

43:47

uptake are going to, uh, sort of, or sort of a

43:50

dynamic readout of changes in uptake can tell us

43:52

how a patient is doing, when they're going to need

43:54

to change therapy, how long they're going to live.

43:58

Uh, AI is certainly going to help us with this.

44:00

We, we have to really mine, mine out these,

44:02

uh, uh, these studies and do larger studies

44:05

that are really going to help us with this.

44:08

And we'll talk about that.

44:10

Uh, and PSA PET has the opportunity to help us in

44:15

places where, uh, where something like a PSA may fall apart.

44:20

So patients will eventually develop

44:21

what may be PSA-non-secreting tumors.

44:23

They may still have PSMA expression, and there are

44:26

difficult situations like bipolar androgen therapy,

44:29

where patients are androgen deprived, but every

44:31

six weeks get supraphysiologic testosterone.

44:33

Well, their PSAs, of course, spike when they get that

44:35

testosterone, but does that mean they're progressing, or

44:38

are, or are we just maintaining sort of the sensitivity of

44:40

their, their disease for, for, for androgen signaling,

44:45

uh, for, for things that target androgen signaling.

44:49

And turns out, if you have any lesions, you are progressing,

44:51

then eventually that will manifest on CT and bone scan.

44:54

But you can tell it earlier on PSMA, so that early time

44:58

point PSMA, again, may be sort of a, a real biomarker

45:02

for, for whether the patient’s doing well on the disease.

45:04

Uh, well on that

45:05

that given therapy or not, and

45:07

whether they need to change therapy.

45:09

With that, I’ll, I’ll kind of wrap up, and I, I think

45:11

I’m about on track to, to have, uh, some good time

45:14

here to answer some questions, but, um, I would

45:16

just like to, to wrap up by saying there are already

45:18

multiple indications for diagnostic PSMA-based imaging.

45:21

We’ve talked about those, but there are interpretive pitfalls

45:23

that suggest a need for a structured reporting approach.

45:26

Doesn’t have to be PSMA-RADS, and that’s my bias,

45:28

but, uh, uh, PSMA-RADS, PROMISE, E-PSMA.

45:32

Uh, I think, uh, all of us should at least think

45:34

about these, uh, think about these possibilities.

45:36

And then, uh, we’re just starting to understand

45:38

PSMA-targeted PET findings as imaging biomarkers,

45:40

and we have a lot of work left to do, uh, uh,

45:43

to really, uh, to really make that happen.

45:46

Uh, with that, I’m going to, uh, uh, stop sharing.

45:49

Uh, I think there’s maybe some questions and some Q and

45:52

A, and I will, uh, I’m going to start, uh, taking those

45:55

maybe, uh, maybe, uh, uh, first come, uh, first come,

46:00

first serve, and, uh, hopefully I’ll, I’ll get as many

46:02

of these answered in the next, uh, uh, 15 minutes or so.

46:06

So looks like the, the first question

46:08

was, uh, what’s the threshold

46:09

PSA after prostatectomy can reliably

46:12

detect recurrent or metastatic disease?

46:14

Um, so this is, uh, this is a, a controversial

46:19

topic, and, but I think a, a very good question.

46:23

So certainly, if you wait longer and this,

46:25

PSA goes up, your, uh, uh, your chance of

46:28

the scan being positive is going to go up.

46:31

But, of course, the chance that the patient also

46:33

has disseminated disease and isn’t going to

46:35

respond to salvage radiotherapy also goes up.

46:38

So what, uh, what I generally tell our clinicians and,

46:42

uh, and our, our clinical practice pattern has been

46:45

that if the clinician is going to pull the trigger on

46:50

salvage therapy, they should get the PSMA scan first.

46:54

Nothing’s worse than, say, doing

46:55

salvage radiation for a patient,

46:57

and then their PSA keeps climbing, and then eventually

46:59

we get a PSMA scan, and sure enough, they had, you

47:01

know, a lymph node that was outside of the salvage field.

47:04

So no matter what the PSA is, if salvage radiation

47:07

therapy is going to be the next step, probably

47:09

need to get the PSMA scan first, uh, uh, again,

47:12

before, uh, just not to have any regrets later.

47:15

It doesn’t mean that the scan’s going to be positive.

47:17

We may not see the site of disease, but we at least

47:20

know that we weren’t sitting on sort of a deal-breaker

47:24

lesion that we just missed because we didn’t image.

47:28

I would say insurances can have problems with, with

47:31

imaging at PSAs below 0.2 because all of the data

47:33

we have is at 0.2 and higher post-prostatectomy.

47:36

So it is, uh, uh, it isn’t clear exactly how

47:40

insurers are going to cover that, but, uh, but from

47:42

a scientific standpoint, I think we’re justified

47:44

to scan patients, uh, so long as it’s before

47:48

before the next step in their therapy.

47:50

And then, uh, the next question, uh, can PSMA PET

47:52

replace MRI for making a diagnosis of local staging?

47:56

Uh, I would say they’re, they’re complementary.

47:58

Um, it’s hard to see EPE, for instance, on, uh, on PSMA PET.

48:03

It is, uh, relatively, uh, you know, it’s

48:07

at least possible to see it on MRI. I think PSMA

48:10

PET is very good for seminal vesicle invasion.

48:13

That’s a little anecdotal, but it just seems

48:15

like it’s really good for seminal vesicle invasion.

48:17

Uh, but what PSMA PET really gets you is

48:20

um, is the, uh, sort of local-regional staging,

48:24

the, the pelvic lymph nodes that may be too small

48:27

to call on MRI. Uh, and then, of course, it gives

48:29

you a, a wider field of view, systemic staging.

48:32

In an ideal world, I suppose we’d all have

48:34

uh, really nice, uh, PET-MRs. Um, but, uh, but

48:38

not every place does, and not every place will.

48:40

Uh, so I, I, I do view the modalities as

48:43

complementary, and I, I think it makes a lot of sense

48:45

for patients to, to get both if, if that’s possible.

48:49

Uh, let’s see.

48:49

Do, uh, do you try to do initial

48:51

staging PSMA PET before giving ADT?

48:54

Uh, again, you know, I guess if the patient’s

48:57

going to get systemic therapy no matter what, the

49:00

PSMA PET may not add a ton of value.

49:03

I, I think the real value of PSMA PET is trying to find

49:05

those patients who have low-volume, uh, metastatic disease

49:09

or local-regionally recurrent disease that, that may

49:12

have sort of salvage options like SBRT. Uh, but if, uh,

49:18

uh, to really get an idea of a patient’s extent of

49:21

disease, I do believe in sort of PSMA PET before

49:24

initiating systemic therapy, just because the PSMA PET’s

49:27

going to be hard to interpret after the systemic therapy.

49:30

As, as you sort of saw in a couple of examples,

49:32

uh, the lesions can do all sorts of things.

49:35

So it is, it's just very hard to interpret PSMA

49:39

PET imaging right after the initiation of an ADT.

49:41

So, so I, I'd certainly prefer to have it, uh, uh,

49:44

early, uh, before than, than just have it after.

49:48

Uh, is there a way to compare the FDG and

49:50

PSA scans for progression response to a

49:52

ADT if a patient has undergone FDG prior?

49:55

Yeah.

49:55

Uh, you know, if a patient's disease is FDG

49:58

avid and, and many castration-resistant patients

50:01

will actually have a fair bit of FDG avidity.

50:03

Um.

50:04

Then, um, then I, uh, I think you can follow

50:08

them by FDG, and that's probably a decent way to

50:11

follow them, uh, because their disease is, is

50:14

going to be from that point on glycolytically active.

50:17

And so if you, if you see that loss of glycolytic activity,

50:20

I, I think that's actually a good way to follow them.

50:22

Uh, a combination of PSMA and FDG. Unfortunately,

50:25

mud muddies the waters quite a bit, so I, uh,

50:29

I'm not, uh, I'm not sure exactly how to sort

50:32

of take the, the combination of the two of them.

50:35

Like if you saw sort of new lesions on PSMA that, that

50:38

weren't FDG avid, you know, would you believe them?

50:40

I, I guess I would, but, um, but it is, uh, it is certainly

50:45

hard to follow with both modalities, and we don't have

50:47

great data on following with both, with both modalities.

50:50

So if they, if they have avidity on one,

50:52

I would probably try to stick to that one.

50:54

Uh, and, and, uh, and follow, follow them along those lines.

50:59

All right.

50:59

Uh, use of PSA in initial evaluation of

51:01

patients found on screen to have elevated

51:03

PSA and may not do well with MRI screening.

51:06

Yeah, so, so there's a little bit of data on this.

51:08

There's a, a, a trial called, the PRIMARY trial from

51:11

Australia uses Gallium PSMA-11, and they reported

51:16

that, uh, that PSA was actually a little bit better

51:17

than MRI at finding clinically significant disease.

51:21

Uh, I can tell you my anecdotal experience is

51:23

that they almost always find the same thing

51:25

if there's clinically significant disease.

51:27

I don't, I don't know that I find PSA

51:29

upfront as, as definitively better.

51:32

Uh, although again, they only sort of publish literature

51:34

would say that, that it's at least a little better

51:37

in terms of patients that can't get MRI screening.

51:40

I, I think the, the,

51:42

the big holdup would be, would be insurance.

51:45

Uh, I think insurance is, is now pretty cool

51:47

with, with MRI as sort of a secondary screening

51:50

modality for patients with elevated PSA.

51:52

Uh, but there are no guidelines that recommend

51:54

PSA PET in that context, and in a patient who,

51:57

in fact, the guidelines specifically say PSA

52:00

PET is only for men with prostate cancer.

52:03

So, uh, an insurance company would have a pretty

52:05

solid foundation to stand on to say that, uh,

52:08

this patient doesn't have confirmed prostate

52:10

cancer, or we're not going to pay for their PSMA PET.

52:12

However, if, if insurance is okay with it, then, uh,

52:16

I think it can be a very useful adjunct to MRI.

52:19

MRI is probably still the gold standard in

52:21

that space, at least in my mind, uh, based

52:23

on the, the level of data that's out there.

52:24

But, uh, but it is, uh, for a patient that

52:27

maybe can't get an MRI, uh, uh, if you

52:30

can get a PSMA PET, it can be useful.

52:33

All right.

52:34

Uh, do I have experience with Pluvicto, and

52:36

what I consider it equivalent to Pylarify?

52:39

Uh, so, uh, yes and yes.

52:42

So I, uh, I, I've, uh, read, I don't

52:46

know, probably 80 or so Pluvicto scans.

52:49

Uh, they, uh, were done on various trials and whatnot.

52:52

And, uh, my, uh, my experience with it is pretty

52:56

much exactly as the companies describe it, or, or

52:59

at least, uh, uh, somewhat as they describe it.

53:02

Uh, there is less urinary excretion than Pylarify.

53:04

So I do think you potentially get a little

53:07

bit of an improved, uh, look at the pelvis.

53:11

Uh, there's probably a little bit

53:13

more false-positive bone uptake.

53:15

Again, that's anecdotal.

53:16

I don't really have data on that, but, uh,

53:18

but my, my observation was I would sometimes

53:20

see things that I was sure were benign bone

53:22

lesions that seemed to have, uh, PSMA uptake.

53:25

Uh, of course I don't have head-to-head

53:27

necessarily PSMA and Pylarify, uh, uh, scans

53:30

in, in a, any significant number of patients.

53:33

But, uh, but yeah, my, uh, uh, my impression from

53:37

the clinical trials that were done is that they are

53:39

statistically equivalent, and we should, uh, from a

53:41

guidelines perspective and, and use of them in, in different

53:44

indications, uh, we should be comfortable with either one.

53:48

Uh, can you share your protocol for PSA imaging for Clarify?

53:51

Uh, do you think that imaging at two hours is beneficial?

53:54

Uh, another great question.

53:55

I think that, uh, I think that there are

53:59

occasionally lesions that show up at two

54:00

hours that don't show up at one hour.

54:02

And it is possible that, uh, it is possible that that

54:07

may make a difference in, in one patient or another.

54:09

Um, I'm not, you know, you, you, you never sort of

54:12

know, there's not necessarily like a big prospective

54:14

file that, that has shown that. I, I would, I would note

54:18

that two hours is outside of the FDA-approved label

54:22

for imaging with PL. Uh, I believe the PPL label is

54:26

something like 45 to 90 minutes or something like that.

54:30

I don't think it goes out to two hours.

54:31

So it would potentially be off-label to use

54:33

it at two hours, which isn't a big deal,

54:34

just, just something to be aware of. Uh.

54:38

What, what I think is maybe the more practical consideration

54:40

is that, uh, at two hours you've used an uptake room for two

54:45

hours, and that may not work with your PET center workflow

54:48

because everybody's getting imaged with FDG at one hour.

54:51

So, uh, so we use, we use one hour because it fits in

54:54

with our FDG workflow, and I, I, I think it's an uncommon

54:58

patient, uh, that's going to benefit from imaging at two hours.

55:01

If you do have the option though, there, there may be

55:03

occasional patients for whom that would be beneficial.

55:06

All right.

55:07

Uh, if, uh, a PSMA scan is negative in patients

55:09

with rising PSA following initial treatment,

55:13

have you found FDG scans to be of any value?

55:15

Uh, no.

55:17

I, I think FDG only really becomes of value when

55:21

you've got relatively advanced castration-resistant disease.

55:24

There are certainly FDG-positive tumors that occur

55:27

before that, but they're relatively uncommon.

55:30

And, uh, you know, I, I suppose the question

55:32

could also be asked, is Axumin useful in that case?

55:36

I, I tend to think not either.

55:39

Uh, because I think most of the negative scans

55:41

in PSMA, in, in that patient population are

55:43

because of volume of disease, and, and there's

55:46

no molecular imaging that's going to find them.

55:48

So, so the question is whether a negative scan

55:50

then triggers salvage radiation therapy still,

55:53

or what we should do with that negative scan.

55:55

But, uh, I think right now, and, and at the guidelines level,

56:00

uh, data would, would, I think, support this statement

56:02

that, uh, PSMA is the most sensitive modality that we

56:06

have for the biochemical recurrence setting, and that

56:08

there's no real reason to believe that other modalities

56:11

are, are going to be reasonable problem-solving tools.

56:14

Um, the one ex, you know, I take,

56:17

there's one exception, I think, to that.

56:18

In a patient who may have a local recurrence,

56:20

maybe they have a positive margin at surgery.

56:23

Uh, dynamic contrast-enhanced pelvic MR is a great tool.

56:26

So, uh, so if you don't see anything on PSMA

56:28

in that context, uh, a, uh, uh, a multiparametric

56:32

MR is, is maybe still a great option.

56:35

All right.

56:36

Uh, if there is an equivocal transitional zone finding, is

56:39

PSMA PET useful to decide biopsy versus active surveillance?

56:43

Hmm.

56:47

Yeah, I, I, I'm not aware of any data that would suggest

56:49

that there, there's, uh, there is data that PSMA PET

56:53

can make you more confident kind of in reading the MRI.

56:57

So I, uh, so if you, if you did, yeah, I, I can certainly

57:01

imagine a case where you've got something on MR. It's PI-

57:03

RADS three, you know, what, what do you really do with that?

57:06

It's blazing hot on PET that you go ahead and biopsy

57:09

that, uh, pattern-forward disease seems to be important

57:12

for PSMA uptake, uh, at least somewhat, but, um, but,

57:17

but again, yeah, not, not, not aware of prospective data.

57:19

So, uh, so it would, uh, again, I think, you know,

57:23

insurance issues may, may sort of come into play,

57:25

but, uh, but it is, it would be, it would be a

57:27

reasonable step from a scientific standpoint.

57:31

Uh, let's see, anonymous attendee, uh, do you, uh, do you

57:34

recommend stopping ADT for some weeks before doing PSMA PET?

57:37

Uh, we don't, whatever the patient's on, uh,

57:39

we figure that's sort of the state of their

57:41

disease, and, and we should go with that.

57:43

So if, uh, if they're on ADT and they've got a rising

57:47

PSA, that would indicate they have viable disease that is

57:51

progressing. It's likely to be PSMA avid, and we should

57:54

go ahead and image them and, and take a look if, uh,

57:56

if, if it's sort of otherwise reasonable to image them.

57:59

So, so, yeah, we don't, we don't really try to change,

58:02

uh, uh, change anything that the clinicians are going to do.

58:05

Again, I think it's not always a, uh, not always a great

58:09

idea to, to try to scan them right after initiation

58:12

of ADT because weird things can be going on, but,

58:15

uh, but we don't, we don't stop ADT generally.

58:19

Okay.

58:19

So, uh, PI three peripheral is in a lesion.

58:21

Is PSMA useful to decide biopsy or not?

58:23

Uh, again, uh, again, I think it, it could be helpful,

58:27

uh, which again, is, uh, uh, extrapolating from, from

58:31

data that isn't exactly in this, uh, in this space.

58:35

Uh, I think it could be helpful.

58:37

Uh, but I would, uh, uh.

58:40

I would, and, you know, and if the patient has, has known

58:44

prostate cancer, uh, even if it's, uh, well, you know,

58:47

I guess the problem is anybody that's on surveillance

58:49

is going to have at least favorable intermediate risk.

58:52

And the guidelines don't recommend PSMA PET in that context.

58:55

So, again, insurers may have a problem with that.

58:57

They, they, I, I don't, I don't, I don't always know

59:00

how nuanced they're getting with things, but, uh, but

59:03

I can imagine insurance being, being an issue there.

59:06

And then, uh.

59:07

Uh, I maybe, uh, maybe have time for, for one more question

59:09

here, and I think there's about one more question here.

59:11

So, uh, are, are we administering Pluvicto, do you think?

59:14

Metastatic history?

59:15

The men, this, men who refuse chemotherapy

59:17

could be considered for therapy.

59:19

Um, so the prob the real problem there is that.

59:23

The, uh, data that shows that men who are not post-

59:26

chemotherapy, uh, and their benefit of Pluvicto hasn't

59:29

been published yet, and so I, I don't believe an

59:32

insurance company is going to pay for a $300,000

59:34

course of, of Pluvicto, uh, on, on that basis.

59:38

It's really the FDA label only for post-chemotherapy men.

59:41

So while I think that, that we're, we're going to move into

59:45

an era very soon where we're giving Pluvicto pre-chemotherapy or

59:48

men who don't want chemotherapy, uh, we're not quite there yet.

59:51

So, so we are only giving Pluvicto right now to

59:54

men who have or have progressed on chemotherapy.

59:57

Okay.

59:57

And then, yeah, one last question.

59:58

So I, I said one, one more, but we'll, we'll do this one.

60:00

Um, do you handle focal intense

60:02

PSMA uptake without CT correlate?

60:04

So it depends on where it is, if it's in bone.

60:07

Um, I look really hard for a CT correlate because

60:10

sometimes the CT correlates are, are very, uh, very subtle.

60:13

They may be like, you know, something that may be minimally

60:16

expansile that might be a good fit for fibrous dysplasia.

60:20

Uh, if, uh.

60:22

So, so in, in bone, I think it can be tricky, but if

60:25

it's really intense, we do have to take it seriously.

60:28

Uh, if it's outside of bone, if it's in soft tissue,

60:31

I guess I'm, I, it, it would be very case dependent.

60:33

Uh, I, uh, uh, if it was truly out in the middle

60:36

of nowhere and there's no lymph node around,

60:39

I, I wouldn't be sure what to think of that.

60:41

Uh, and other organ systems, it, it may require

60:43

additional workup, like if it's in the liver,

60:45

uh, that, that could be an HCC or something.

60:47

We may have to go hunt that down a little bit.

60:49

Uh, but, uh, uh, but yeah, it would, it would

60:52

sort of, it would sort of depend a little bit.

60:54

Uh, and then if it's in the prostate, uh,

60:56

it's almost certainly prostate cancer.

60:57

And, uh, uh, and that's presumably, I guess,

61:00

why the patient would be being imaged.

61:02

Um, and with that, I do apologize.

61:04

I, I, I am going to have to run.

61:06

I, I know there's, uh, there's probably other

61:08

questions that'll trickle in, but, uh, it's

61:10

really been a pleasure, uh, uh, joining you today.

61:12

And I hope this was, uh, I hope this was a useful, uh,

61:16

a useful, uh, uh, talk, and, uh, really

61:18

thank you all for your attention.

61:20

Uh, thank you for some great questions, and, uh, uh,

61:23

hope everybody has, has fun reading these scans.

61:25

Thank you again.

61:25

Thank you so much, Dr. Rowe.

61:25

1371 01:01:27,170 --> 01:01:27,950 Thank you so much.

61:28

Um, you can access the recording of today's

61:30

conference and all our previous noon conferences

61:33

by creating a free MRI Online account.

61:36

Be sure to join us next week for a noon conference

61:38

double feature, uh, starting on Tuesday,

61:40

July 25th, at 12:00 PM Eastern, featuring Dr.

61:44

Steven J. Pomerance for a lecture on wrist MRI, and on

61:47

Thursday, July 27th, at 12:00 PM Eastern, Dr. Deborah

61:51

Baumgarten for a case-based review of renal pathology.

61:55

You can register for this free

61:56

lecture at mrionline.com and follow us on social

62:00

media for updates on future noon conferences.

62:02

Thanks again, and have a great day.