Trends in Alzheimer Imaging & Management

In this discussion Dr Tanenbaum, Suzie Bash, MD, a medical director at RadNet, and Petrice Cogswell, MD, PhD, associate professor of radiology at The Mayo Clinic, Rochester, Minnesota, discussed opportunities and responsibilities facing the neuroradiology community.

Creators and Guests

Host
Lawrence Tanenbaum, MD, FACR
Dr. Tanenbaum is a long term collaborator with the medical imaging industry, with interests in developing applications of contrast agents, MR, CT and advanced rendering in the clinical practice of medicine focusing on efficiency, radiation dose appropriateness and physiologic imaging. Dr. Tanenbaum has authored approximately 100 scholarly and peer reviewed articles, continues to chair educational and academic meetings and has delivered over 1500 invited lectures around the world.
Guest
Petrice Cogswell, MD
Guest
Suzie Bash, MD
Neuroradiologist. Interested in Artificial Intelligence (AI) applications in advanced Neuroimaging, MRI, FDG brain PET/CT, Amyloid PET, Dementia, Multiple Sclerosis (MS), Quantitative Volumetric MRI (QMRI), Deep Learning Image Reconstruction (DLR), Clinical Research, DL, ML, Medicine, CT, Digital Imaging, Neuroradiology, and music.

What is Trends in Alzheimer Imaging & Management?

Moderated by Lawrence Tanenbaum, MD, FACR, Applied Radiology hosted a series of conversations with experts in Alzheimer disease imaging and management at the 2024 Radiological Society of North America Annual Meeting and Scientific Exhibition.

Larry Tanenbaum, MD:

Hello. I'm Larry Tannenbaum, and welcome to the floor of RSNA 2024. I'm thrilled to have the opportunity to welcome you all here to this applied radiology roundtable series on the latest trends in Alzheimer's imaging and management. For this module, a call to action, obligations for neuroradiologists and the enterprise, I'm joined here today by doctor Patrice Cogwell, who's an associate professor of radiology at the Mayo Clinic, and doctor Susie Basch, who is from RadNet and who runs he was the national lead for all of, Alzheimer's imaging and dementia imaging, as well as quantitative imaging. Doctor Basch.

Suzie Bash, MD:

Thanks for having us.

Larry Tanenbaum, MD:

Alright. So let's get into it. There's a lot to discuss today. This is one of my most I one of the areas that I have the greatest passion about in this in this area of dementia imaging. My my co mod my co panelists are, you know, wondering what happens when I have too much passion, but let's let's just move on.

Larry Tanenbaum, MD:

Can can one of you just sort of give me a sense where we're gonna get started? Talk to me about the disease modifying therapies. What are they, you know, and the rest of the background. Who's gonna take that one? Great.

Petrice Cogswell, MD:

I'll go ahead with that one. So the disease modifying therapies for Alzheimer's disease, the general, general field of therapies are, immunotherapies targeting beta amyloid. Okay. We can remove amyloid from the brain. There's currently two that are approved by the FDA, and those are laconumab and denanumab.

Petrice Cogswell, MD:

And those agents have shown to provide about a 20% slowing of cognitive decline and a marked removal of amyloid from the brain. So patients who have, amyloid in their brain before starting therapy, will have amyloid removed to the point that on a PET scan, they would be considered amyloid negative.

Larry Tanenbaum, MD:

So it's all based on the theory that amyloid is a causative agent here and removing them improves therapy. Is there any stratification as to how weight patients will do, say, based on how much amyloid or how much tau they've got?

Petrice Cogswell, MD:

So the that is a good question. And that is

Larry Tanenbaum, MD:

I have a few of those.

Petrice Cogswell, MD:

That has that has yet, yet to be answered. So in terms of the, the amount of, well, there's some data showing so far that patients who are more advanced in the disease process, so more tau in their brain don't do as well. So that targeting patients earlier, those patients will have the best response to therapy in terms of the slowing of cognitive decline.

Suzie Bash, MD:

And actually, when we looked when the CLARITY AD trial, when they looked at their data, they found exactly that. So when they did sort of a sub study analysis of the low tau group patients, they had much better clinical outcomes than the twenty percent that you just reported. So actually, seventy six percent of those patients had no cognitive decline at eighteen months, and sixty percent actually had cognitive improvement at eighteen months when you catch them in the earliest stage of disease. In other words, the low tau sphinct.

Larry Tanenbaum, MD:

Alright. That's great. So, the therapy sound wonderful. Are there any risks associated with these, therapies?

Suzie Bash, MD:

So, unfortunately, with monoclonal antibodies that target beta amyloid plaque, you have the risk of amyloid related imaging abnormalities or ARIA. So there's ARIA e, which is edema or focal effusions, and ARIA h, which is micro hemorrhage or superficial sclerosis. Now the majority of patients won't even know that they develop ARIA. So for example, with Lekembe, only three percent of patients, have symptomatic ARIA. And for it's a little bit higher for denanumab.

Suzie Bash, MD:

It's, for Kinsenla, it's a six percent. So again, most patients don't know, but it's still critical to be able to identify the ARIA when it does occur so that the infusion can be held if it's moderate or severe degree radiographically. Because what you don't wanna do is dose a patient on top of, you know, a moderate or severe ARIA and then potentially have it get worse. Now ARIA can get worse on its own anyway, but ARIA e, the good news is is that it essentially almost always resolves by one year and then the ARIA h will become stable. And in my experience, these patients are very they they notice the benefit of the drug and are actually anxious to restart it once their ARIA has resolved.

Suzie Bash, MD:

But again, for neuroradiologists, it's very, very critical that we, find the area, grade it correctly, and communicate that back with the referring neurologist so that they can hold the infusion, if they need to.

Larry Tanenbaum, MD:

So it's an interesting thing because, you know, there are the two approved drugs that we mentioned, and it did does seem to be a trend that when the efficacy goes up, the complication rate goes up. It's really kind of exciting to hear the recent news that when you adjust the dosing regimen for Kesemla, the the complication rate sort of settled down to about where it was with Lekambi, which is really exciting. Because there are more drugs coming. They do promise perhaps to be more effective, but they also promise to be have higher risks. So it'll be interesting to see how that plays out.

Larry Tanenbaum, MD:

It's good stuff. So does any either one of you wanna talk about why we get edema and hemorrhage when we mobilize amyloid?

Petrice Cogswell, MD:

So the edema and hemorrhage, are thoughts of of ARIA, are thought to be, the similar pathophysiology to cerebral amyloid angiopathy related inflammation. So, the exact etiology of ARIA is unknown, but, they're supposed thought to be multiple factors. So related to the removal of amyloid from the vessel wall, as well as the parenchyma, with the amyloid removing therapies, and that can create, so let's say, leaky vessels. It is also thought that as the amyloid is removed from the parenchyma, some of that may deposit back into the vessel wall and, hence, a a worsening of cerebral amyloid angiopathy.

Larry Tanenbaum, MD:

Because early on, the predominant concentration is in that perivascular space, the periarterial space. So I hadn't really thought about the fact that after you clear it, it's gonna it's gonna fill in where the other one goes and, yeah, and attack a vessel that may very well be already a flame and leaky. I think that's really a good way to describe that and to help understand that. Alright. So let's let's go one step further.

Larry Tanenbaum, MD:

So we talked about the therapies. We talked about the risks. And I wanna expand more on what Aria is in a moment because that's really where the focus of this session is. But let's start out with you're doing the MR. You're going to give these therapies, which may cause these the amount of hemorrhage.

Larry Tanenbaum, MD:

What are we looking for? Either one of you or both of you.

Suzie Bash, MD:

Oh, I can start. So yeah. So a baseline MRI is required prior to starting, therapy. And the key components of the baseline, MRI interpretation is identifying any baseline microhemorrhages. We know that patients with Alzheimer's disease are more likely to get cerebral amyloid angiopathy.

Suzie Bash, MD:

We don't wanna start a patient on therapy and miss those micro hemorrhages. And it's not just about finding a few, we actually need the actual count. It's extremely important to know the baseline count of microhemorrhages.

Larry Tanenbaum, MD:

Why is that important?

Suzie Bash, MD:

Because what we don't wanna do is neglect to mention baseline microhemorrhages, and then have the patient come back and get an ARIA interpretation or surveillance study. And now we're mentioning all these micro hemorrhages and we don't wanna attribute that to the drugs if it was already present in baseline.

Larry Tanenbaum, MD:

Let me just take you back one step, you know, for a simple minded guy. Why do I care how many hemorrhages are on the baseline exam?

Suzie Bash, MD:

Well, because the neurologist may not feel that the patient is a good candidate for therapy if they have a lot of baseline microhemorrhages. Now the

Larry Tanenbaum, MD:

clinical macrohemorrhages too. Right? If they have a a diseased brain that maybe has had a lot of strokes or prior prior, you know, hemorrhages You're

Petrice Cogswell, MD:

in that. Those right. Those are findings that are that increase the patient's risk for ARIA. Right. And that right.

Petrice Cogswell, MD:

It That's a very important piece. Right? Microhemorrhages or cirrhosis would, right, give it puts the patient at increased risk of having both more microhemorrhages and cirrhosis as well as edema with Yeah.

Larry Tanenbaum, MD:

Because that part doesn't get as much coverage because we're always focused on ARIA. But I think it's important for the neuroradiologist to understand what the criteria are in those what we're looking for in those circumstances, which I think is very, very important. Talk to me about the surveillance examinations, what we're looking for, how important is it to you know, how and how important is it for us to do our job properly?

Petrice Cogswell, MD:

So on the surveillance MRIs where we're looking for ARIA, so we're looking for findings of new edema, a new subtle effusion, new microhemorrhages and cirrhosis since starting therapy. And to do that, we need to have a high quality brain MRI or preferentially preferred at or performed at three Tesla. And we need to have the appropriate sequences. So we need to have a t two flare to look for RAE, and a GRE or SWI to look for the RAE microhemorrhages and cirrhosis. And to do this well, we ideally perform this in a consistent perform the imaging in a consistent fashion across the serial exams in a patient.

Larry Tanenbaum, MD:

Great. So I I I want there's a few things I wanna, pick up on. You know, as you know, I was once an administrator for a large imaging organization. We had about a third three t scanners, but not all of them, you know, with three t. So, you know, while it might be nice to prefer a three t, it's kinda difficult to accomplish in the community.

Larry Tanenbaum, MD:

And, also, you know, while it might be great to be seeking on a 1.5 or a three t, sometimes you gotta go where the patients can go, and sometimes that means a 1.2 Tesla scanner. And in some settings, when you start talking about going to unusual settings, it might mean a point five Tesla scanner or a point zero six Tesla scanner, which will be very interesting to topic to see where that actually goes. So you talked about what we have to detect, but I don't understand I'm playing like I don't understand. I don't understand why it's important to count and detect all of these lesions. Are there some rules that I need to know about About how whether therapy would be, paused or continued?

Suzie Bash, MD:

Yeah. So it it is important to have an accurate count. But getting back to what you were saying before as well, the protocol is extremely important. For example, we were talking here about a GRE or an SWI. Again, the parameters standardization is really ideal when it comes to screening these patients.

Suzie Bash, MD:

In an ideal world, they come back to the exact same scanner or the exact same strength, but that doesn't happen in reality as you were just mentioning. But what is very important is that we make sure our protocol parameters are in place. So, you know, we we need to make sure that the t e is correct. So ideally on a 1.5 scanner, we want that t e around 30. On a three Tesla, around 20.

Suzie Bash, MD:

If the t e is set too low, you will miss the micro hemorrhages. And so, it anticipating operational issues that could happen, if a patient came in and they scan they were scanned on a one five, the TE was only five and then microhemorrhages were not caught on the baseline MRI. Now the patient comes back, they're on a different scanner, maybe a three t, and now the TE is set correctly and you've identified now six new microhemorrhages. The problem is is is that ARIA or was it due to improper technique at the baseline MRI? So again, that's why it's very important for imaging enterprises to have, to really think about this, have set protocols in place, and scan their dementia protocol the same way each time and preferably with a three d t one input as well so it's quant capable.

Larry Tanenbaum, MD:

So what I was trying to tease out of my panelists, so now I have to do myself, is is that there are criteria whether therapy can be continued or has to be paused, and it's based on the grading and severity of these findings. So it's really important that we do this properly. We be tuned in to the appropriate inflection points when therapy is changed and communicate that, communicate that effectively. You mentioned, doctor Basch mentioned that, the the if you're going to use gradient echo, which is sort of the baseline way we've detected hemorrhages for generations, that the parameters are really important. And I see that across the nation, that that's one of the areas where we fail.

Larry Tanenbaum, MD:

Even when I was in charge of the imaging enterprise, we failed. So it's very, very important to pay attention to that. But as long as I have zealots here at the table with me, let's do the quick discussion on GRE versus SWI. What's your opinion? And then, Patrice, I'll let you go ahead.

Suzie Bash, MD:

Yeah. So, at at RadNet, we've decided to use the GRE for all of our dementia protocol, cases. And that's because that's what they used in the trial and that's what the ARIA grading scale was based off of. So we know that the SWI is more sensitive for picking up these micro hemorrhages, but we don't encourage use. We encourage just the GRE if possible.

Suzie Bash, MD:

Now if the referring, doctor wants an SWI, we'll do that as an add on to the GRE, but we always do the GRE so that the grading is consistent and matches the scale. Truthfully, the grading system would need to change if if it was for SWI. So it was based on GRE, so that's why we do that. And the problem with doing both the GRE and an SWI is maybe the SWI picks up a couple more micro hemorrhages because it's a more sensitive technique than the GRE, and then it kinda muddies the waters with our reports, you know. So we're saying, okay.

Suzie Bash, MD:

This is the number of microhemorrhages I see on GRE. This is the radiographic ARIA grading, but this is what the SWI found. It just it's just not nearly as clean of a report, And that's why at RadNet, we've decided to do that. But, again, consistency is really what's paramount. So what's been your experience?

Petrice Cogswell, MD:

So we are doing right now an SWI and a GRE for both the baseline enrollment scan as well as ARIA monitoring. Some of the radiologists have asked, like, so then what do I report? Do I say, you know, do I report on both of them? And we've said, report on what you what is the true number of micro hemorrhages. And so I think and then it has been interesting to see, you know, how the micro hemorrhages appear on the SWI versus GRE.

Petrice Cogswell, MD:

And sometimes there are, you know, rare cases where on a GRE you can more definitively call this or that is a micro hemorrhage, but right, the SWI. Vessel.

Larry Tanenbaum, MD:

Yeah. And that that's the part we don't usually think about because it's a little more challenging to interpret an SWI. So I I I'm I'm very interested in your perspective. So while in any given visit, you can establish we know that SWI is more sensitive, maybe a bit less specific. You can use whatever baseline you like, except that when you get incident changes, changes in the interval, and the rules and guidelines that we have as they exist are based on the lower sensitivity GRE.

Larry Tanenbaum, MD:

Let's say for a theory, you have no micro hemorrhages on the GRE incident from the last exam, but now you've got three micro hemorrhages incident seen only on the SWI. How would you react to that? And let's say that wasn't three. Let's say it was five. Get over the threshold for you're gonna get you may wanna change.

Larry Tanenbaum, MD:

You can see five new ones on the SWI, but only two maybe on the GRE. How would you deal with that?

Suzie Bash, MD:

How would you grade the REI?

Petrice Cogswell, MD:

Right. Well, I think so our our neurologists are very on board with a conservative approach. So they would want us to they would want to over then overreact then underreact to something. And, they are cautious in following up microhemorrhages even if it were two or three to follow those and make sure that they are stable. So they, right now want to go off the the SWI findings or whatever it is that you think are the real, you know, number of microhemorrhages they want to be as as sensitive as possible.

Suzie Bash, MD:

And so would they hold the infusion if the SWI got you over that, five threshold, but the GRE didn't?

Petrice Cogswell, MD:

Yes. They would go. They would, you know, they'd take the whole picture into context, but they would, manage based upon the SWI for the time being. Yeah.

Larry Tanenbaum, MD:

So I only have really time for one more topic, and that is about, interfacing with neurologists and other shareholders in the treatment pathway. Can you talk to me about, you know, the Mayo Clinic is a good place, you know, with an integrated health system to have this question. Multidisciplinary patient management, what do you do when the genetic testing points to one or two alleles, you know, of the e four? Do you do the genetic testing in your institution?

Petrice Cogswell, MD:

Because I've seen in some academic centers it's becoming required. When we first started this, it seemed to be more optional. Talk to us about that. All of the patients that go through our Alzheimer's disease treatment clinic for evaluation, they all get APOE genotyping. And we have a weekly multidisciplinary conference in which we talk about their neuropsychological testing, their imaging biomarkers, and their APOE status is incorporated into their risk benefit discussion.

Petrice Cogswell, MD:

And that is so in our we talk about that at conference, and then they integrate a number of micro hemorrhages and their APOE four status to the biggest risk factors for ARIA in that into their, the risk discussion, and that risk discussion we'll have with the patient.

Larry Tanenbaum, MD:

So two questions. Is there a strict formula that you follow, or do is it has it become gray areas and that kind of the same?

Petrice Cogswell, MD:

There are definitely gray areas. We we're not treating APOE four homozygotes at all until more recently, and and they're taking a a similar cautious approach with that. But they'll take it into context knowing that, you know, four microhemorrhages may still be eligible for therapy, but at greater risk than two microhemorrhages. So again, part of the risk discussion with the patient.

Larry Tanenbaum, MD:

Would you just clarify for me because I'm fairly naive about this. Who is in that multidisciplinary conference? Which disciplines?

Petrice Cogswell, MD:

So it is, our cognitive neurologists, who see all the patients, the neuropsychologists, neuroradiologists, and nuclear medicine physicians. And then there's also social workers and I think other parts of the allied health care team, who help with with the patient care pathway.

Larry Tanenbaum, MD:

Yeah. I think it's wonderful that that this type of multidisciplinary tumor board, so to speak, is exists, and it is guiding therapy in these patients. Well, I think this is as good of a place as any to wrap up our discussion on a call to action, obligations for neuroradiology in the enterprise. I'd like to give great thanks to my panelist, doctor Susie Bash, doctor Patrice Cogswell, for joining me in this roundtable series in the latest trends in Alzheimer's Imaging and Management being hosted by Applied Radiology with sponsorship from Siemens HealthCare and AgoraCare. Thank you very much for joining us.

Suzie Bash, MD:

Thanks for having us.

Petrice Cogswell, MD:

Thank you.