Chapters Transcript Wishful Thinking: What Are the New Treatments in Cognitive Neurology? Course: Neurology Updates 2023: Case by Case And I actually wanted to take a step back, um, because, you know, while treatments in our field are advancing and side effects of those treatments are also advancing and understanding of those uh is critical, um, there's also in parallel the diagnostics because we have to diagnose our patients to be eligible for these new uh medications. And also we have to be able to uh address any side effects. So, uh, uh, where do we stand with diagnostic testing in Alzheimer's disease and related dementias where. Like I said, the advent of these new disease modifying therapies encourages accurate identification of Alzheimer's disease at early stages at the MCI stage, the mild dementia stage. And so, you know, the trends that I see is that there's a, a classic diagnostics are really based on clinical phenotyping, the classification of the dementias has been largely based on symptoms, uh, whether they be cognitive, physical, etc. but I think there's a movement to augment this classification or um add to this classification molecular biomarker. Classification that incorporates information about amyloid tau neurodegeneration, even alpha synuclein inflammation and vascular lesions that all add to the phenotype. And the other part of this is that we now know that phenotype doesn't necessarily dictate pathology, for example, there are atypical forms of Alzheimer's disease that look like aphasia syndromes or have a visual presentations or motoric presentations like corticobbasal syndrome, um, or even ones that look like frontotemporal dementia, that this executive behavioral form of Alzheimer's disease. And the other trends that I will review at the end are towards more cost effective and non-invasive approaches so that hopefully we can increase screening and, and improve access to new therapies as as they're developed. So I wanna start with just a review of neuropsychological testing. I think that a lot of us are really familiar with this. um, it's not an advanced technology, but it's quite helpful in particular situations. So I view that neuropsychological testing is useful to determine the pattern and severity of cognitive impairment and when do I use it? Well, there are a couple of situations. One is where I'm really not sure if someone is cognitively normal or has very early MCI. Because my bedside exam is unclear or perhaps there's a discordance between my exam and what uh a patient is saying or between what a patient is saying and what their family member is saying. Uh, a second situation is where someone is at an earlier stage, and I want to determine uh what the etiology is or if there are multiple ideologies, uh, whether they be neurological or non-neurological. And especially in cases where these other diagnostic approaches that I that I will go through are somewhat equivocal, where you have mixed picture pet imaging or um the CSF biomarkers don't really give a full uh clear picture. So MRI, uh, I think we are also familiar with MRI. We know that we use it to assess for structural lesions such as stroke, uh, tumor, subdural hematoma, um, uh, uh, normal pressure hydrocephalus, right? And of course the, uh, very traditional structural biomarker of Alzheimer's disease, is hippocampal atrophy. Um, which, uh, you know, has been long been been known in the field, um, especially in amnestic presentations, um, but you know, as Doctor Sadowski, uh, reviewed in great detail, we also use MRI to, to, to screen and evaluate for, for uh area, um. I also wanted to review, um, white matter hyperintensities because this is often commented on on MRI reports and we're not exactly sure what to do with it. I mean, of course we feel that, you know, severe confluent white matter hyperintensities may indicate um a vascular dementia and may may indicate a higher risk for for area, um, but I think there's recent evidence of the location of the white matter at hyperintensities can also associate with different processes. So when the figure to the left is delineated, um, sort of a, a distinction between more perivventricular white matter disease and more subcortical involvement. Um, subcortical disease is, is typically associated with, say, diabetes, um, and paraventricular disease, uh, certainly can be driven by hypertension, but it has also been linked to, to Alzheimer's disease, um, and even FTD itself, um, in particular, more, uh, posterior, uh, paraventricular disease, uh, has been associated with AD. Uh, whereas, uh, isolated anterior disease can be found in, in certain FTD syndromes. So just, you know, keep that in mind when, when you do see white matter hyperintensities in, uh, sort of unusual cases or cases where neurodegenerative disease is, is high on the differential. And I don't want to dwell too long on this because we had had an excellent talk on AREIA, but you know, uh, uh, MRI is also used to evaluate for REAE and REAH, which, as you as was just reviewed can occur both spontaneously due to Alzheimer's disease but or due to immunotherapy. It's vasogenic in origin and uh its location can be uh payal or uh leptomeningeal. Um, and typically the, the, um, the, the sequences used are, are flare, um, and, uh, uh, T2 star GRE or or SWI for the, the hemorrhagic component. So the next level testing that we have available to us is uh the functional scans so those come in forms of FDG PET or HMPO specs. So FDG PET is really looking at glucose usage, so hypermetabolism pattern, and HMPOE is uh sort of analogous but looking at hypo perfusion patterns. And you know, from a very basic perspective you're looking at um sort of a a heat map of the brain where uh normally functioning active regions are typically color coded in in red in the processed form and the uh more abnormally functioning regions are are colder colors and uh that pattern is typically aligned with a particular condition. Um, FTGPET, uh, you know, CMS guidelines are for use to distinguish AD versus FTD. It's, uh, moderately superior to SEC. I will use SEC in, uh, patients where I cannot get, uh, FDG PET. It, it, it seems to be a reasonable substitution. Um, the hypometabolism or hypo perfusion patterns can suggest different conditions, a classically a temporal parietal pattern. Of hypometabolism, hyper perfusion will be linked to Alzheimer's disease, whereas a frontal temporal or frontal anterior temporal, uh, pattern will be more associated with FTD. Um, there is this, uh, uh, consideration that occipital involvement in a particular way may also suggest, uh, Parkinson's disease or lewd body disease. So examples of these are sort of shown, um, in the two figures here, so that to the extreme right. Is uh an FDGET of uh a DLB patient, uh, where you get, um, you know, occipital, uh, involvement but different from Alzheimer's disease you get preservation of the of the posterior cingulate, so this is a cingulate island sign. Of of Lewy body disease. So I, I would say if it's if it's there and there's a high pretest probability of of LBD, I think this is helpful. I think if, if the pattern is not there, then I wouldn't necessarily exclude Lewy body, uh, especially if your exam and history is, is suggestive. And of course these are multiple patterns of FDG showing the um temporoparietal pattern in Alzheimer's disease and how different forms of um FTD can appear so uh a semantic variant which is more uh temporal and a behavioral variant which is more uh frontal, um, and a loopenic uh pattern which may be more overlapping with Alzheimer's disease. But you have to remember that these patterns are not unique and I definitely have had several patients where, um, you know, they have frontotemporal patterns on pet, but they ultimately have Alzheimer's disease biomarkers and uh sometimes occipital, there's occipital involvement, but there's no Parkinsonism, no REM sleep behavior disorder, and in fact they have the like a posterior cortical atrophy variant of Alzheimer's disease. So keep in mind that, that, you know, when you see the FDG PET, put the, put in the context of the patient of the history and you may need to get uh subsequent biomarker analysis which I'll, which I'll go over next. Um, sometimes we get mixed or incomplete patterns. I've had several patients where I have only temporal lobe involvement, only occipital, only parietal and then uh it's not clear what this represents. Maybe they have to be scanned again in time. Maybe you do need to get biomarker assessments, uh, maybe it's suggestive of a another condition, for example, uh, temporal lobe epilepsy or, um, uh, an autoimmune encephalitis, for example. Um, I will also have patients who have mixed patterns frontal, temporal and parietal involvement. I tend to feel that if the parietal lobe is involved that would suggest more of a. More of an Alzheimer's disease, uh, process, um, but certainly, Lewy body disease can look like this as well, uh, and less likely, uh, FTD. So now we're in the era of more molecular or bio biomarker based uh uh testing so we now have, um, you know, amyloid PET uh available um it has a uh sensitivity of 89% as a specificity of 87% to distinguish Alzheimer's disease patients from normal controls. also who is quite good at differentiating Alzheimer's disease versus FTD um, to the right is an example of the difference between amyloid PET. Uh, scan of, of someone who is normal versus someone who has Alzheimer's disease. I just wanted to note that even in a normal, uh, patient, you will get tracer uptake in in the white matter. And so when these visual reads are done, they have to take that into account. There are multiple FDA approved tracers, uh, Florbeaben, Flubeta uh Florbetape and uh uh flumetamol, um, they are visually red. As positive or negative and typically positive means that there is significant cortical uptake, basically a ring of of of cortical uptake, um, and, uh, recently CMS did lift uh uh coverage restrictions which allowed local Medicare administrative contractors to determine coverage and this is really important because. It, it really is in our algorithm for example, for uh vetting, uh, patients for for leanumab we really do have to verify that they have amyloid positivity. So these scans don't have to always be done. um I think that if if if the syndrome looks like Alzheimer's disease and if, if you wanna be extra careful and you, you do an FDG PET looks like AD more often than not you can stop there but I, but if they are interested in, in uh uh licaumab. Um, or if the, uh, uh, PET imaging or SPECT imaging is equivocal and it's really important for other reasons to uh determine the AD diagnosis, then, then, you know, amyloid pet can be very useful. Um, but just keep in mind that, that, yes, it's, it's useful to differentiate Alzheimer's disease from um FTD as shown, as shown in the right, right? So, so you have an Alzheimer's disease patient who's amyloid positive, a lot of uptake in the cortex, whereas the um FTD patient will have very minimal cortical uptake. Keep in mind that Lewy body patients, especially older ones, will be amyloid positive and so you cannot really use that to distinguish the two. And really we do not use it in normal uh patients. There's there's no need to order these on uh to, to assess amyloid in normal people at the, at present as unclear predictive power, um, and, and, and one of the additional reasons to to for this and to keep in mind is that amyloid positivity or amyloid itself is basically increases with age. So 10% of people who are age 50 are amyloid positive, 15% at age 60, 20% at age 70, 30% at age 8, and 40% at age 90 are amyloid positive. And um well this actually supports why you need for example tau sometimes as an additional biomarker and of course borderline cases because visual reads require, you know, so much cortical involvement there may be cases where you have isolated uh parts of neocortex involved and depending on the clinical setting, you may need to uh repeat the imaging in time or do CSF analysis. Um, just to allude to the Lewy body example, here's a paper that showed, uh, one Lewy body patient that was clearly amyloid negative, but another Lewy body patient that was, um, amyloid positive. Um, another tool in our arsenal now is CSF measurements of, of Alzheimer's biomarkers, and the pattern that you're looking for is actually low, uh, CSF, a beta 42, especially when ratio to the overall tau, um, so it's inversely proportional to to brain, uh, amyloid. And high PTA, so that's more uh um uh uh directly correlated with uh with brain tau, and that's highly sensitive and specific for AD pathology in the brain. So in research settings the sensitivity is 92% specificity 86%, it is somewhat lower in in clinic settings 83% and 72% respectively. um, there is a thought that CSF amyloid reductions or. Abnormalities may precede accumulation of brain amyloid, although there is some mixed evidence on this and and still something that needs to be sorted out and of course CSF analysis allows for testing for other other causes like autoimmune inflammatory infectious and and neoplastic. Uh, I just wanted to go over what some of these readouts look like. I mean sometimes you just get the numbers, um, you know, oftentimes you get sort of a visual description like this, and I just wanted to go over what you're sort of looking at when you get this. So, um, typically you want to look at it as a, as a 2x2 kind of square with uh P tau as the X axis and the amyloid tau ratio or amyloid tau index as the Y axis. And um so the ATI is amyloid and the PTA is phosphorylated tau and of course this patient here is clearly in the amyloid up in the Alzheimer's box so there's clearly high P tau and a low amyloid tau index below the cutoff. But what about these areas? We do have some patients who were one or the other biomarker falls in this category. And what this really means is it's, it's, it's um it's equivocal so it's unclear if if it's sort of in this gray area here it's unclear if the PTA is positive or negative in brain you just don't know and if the uh ATI is in this range it's it's sort of unclear if amyloid uh is is there in the brain or not and so you have to keep this in mind, keep it in the context of the patient and their um clinical syndrome. So T pet is, uh, there is one tracer that is FDA approved for tangle pathology and Alzheimer's disease, uh, you know, typically it, it is, um, uh, uh, you really want it's really good for, uh, uh, tracer uptake in neocortex, um, that really contributes to the interpretation of, of, of positivity. It's not really covered by insurance currently. And um because it it's, it's better at sort of neocortical uh tau is very good at determining severe uh tau pathology may be less useful in milder cases, uh, it's not to be used for other forms of um teopathy such as CTE. It's, it's not really widely used currently, um, but ideally in the future maybe combined with amyloid PET. So here's an example of, of sort of a control and uh a brain that has a lot of uh uh tau. So I get a question a lot of what's the use of genetic testing and I think that it's good to look at this both in terms of the autosomal dominant and the risk factors. So genetic testing in terms of uh autosomal dominant genes, so less than 1% of, of late onset AD is autosomal dominant, although 8%. Of early onset AD is uh autosomal dominant, meaning it's driven by mutations on the APB PS1 and PS2 genes. Um, these mutations can be hereditary or sporadic. Um, I view these genes as, uh, not terribly useful in, in late onset, uh, Alzheimer's disease, but can aid in the diagnosis of early onset patients, especially the biomarkers aren't clear or not available to be done. Um, but I always do encourage genetic counseling to be done in conjunction with this testing. In terms of FOE 4, which was mentioned before, 11 copy confers a um fourfold risk. Excuse me, of um Of Alzheimer's disease, whereas 2 copies confers a 14-fold increased risk. It is race and ethnicity dependent. So just be careful about how you interpret this across different uh groups, demographic groups. It's really not useful for diagnosis because it's not deterministic, but it is useful. In terms of counseling patients, uh, by lecaumab. And uh AI side effects as was reviewed before. And lastly, I want to end with uh the future. So plasma biomarkers, and I think the bottom line here here is not there yet but soon. um this is sort of a table that I created of all the different uh things that we can test in plasma uh you might be hearing about in the coming years. Um, so there's a beta 4240 you might be hearing about, um, some offerings of those. It's really not terribly, uh, useful. It's, uh, in terms of the research data that that we've seen, it is still inversely proportional, so a lower value will sort of loosely correlate with with brain amyloid. But not very helpful with differential diagnosis seems like total tau also is very non-specific for neurogeneration at large. Um, the most useful one seems to be PTA, either PTA 181 or PTA 217. It seems to correlate well with both amyloid and PTA, uh, deposition in the brain, um, and it may be useful in the future to, to diagnose AD to differentiate from FTD or. Uh, and other diseases, but, um, that that seems to be the most promising, uh, biomarker, uh, that, that may be used in the future. Um, other plasma markers like NFL and GFAP certainly can be complementary. They do represent um specific pathophysiologic processes such as. Uh, aolenol, uh, and other forms of neurodegeneration or, um, inflammation and astroglu activation and, and loose correlations with brain amyloid and so they may be complementary, uh, to some of these other plasma biomarkers in the future. So there are several uh companies that are that are working on these tests. uh you may be reading about this, you may be even getting inquiries, but the bottom line is nothing has really been completely approved uh for clinical use. Um, they vary between a beta 4240 tests sometimes combined with Apple E4. Um, and then there is a, a PTA 18181 test, um, there others being developed and PTA 217 also being developed, um, you will see that some methods use mass spec, some methods use immunoassay, uh, generally mass spec is considered to be, uh, better. But practical considerations moving forward. PTA tends to be the best one, more sensitive and specific to Alzheimer's disease, and then within that PTA 217 better than 181 mass spec appears to be better than immunoassay based, uh, assays. Uh, application to the clinical population we're finding out is, uh, unclear because of the, the, the, the effect of comorbidities on a lot of these cutoffs such as diabetes and renal disease. So I think we have to be particularly careful when once we apply it to a very general population. But, um, you know, when widespread use is accepted, we hope that it can be used to diagnose impaired patients. We don't really know their utility and normal persons and of course there's no uh coverage for these tests at present. Um, so I just wanted to end with some broad take home points. One is MRI remains quite useful for initial assessments of, of non-degenerative pathologies and screening for area. I think we're gonna be, uh, using more and more amyloid PET and CSF amyloid tau analysis so we can actually, um, diagnose AD in the setting of a reasonable pretest probability and move forward with, with, uh, you know, anti-amyloid immunotherapy. And lastly in the coming years look for plasma biomarkers to be increasingly validated and potentially available for clinical use, so I'll stop there and thank you very much. Published December 15, 2023 Created by Related Presenters Arjun Masurkar, MD, PhD View full profile
CME Knowledge Hub
