The director of the Brain Health Observatory at the University of Southern California discussed the potential impact of Alzheimer disease blood tests on reducing clinic wait times and improving the diagnostic process in primary care.
Soeren Mattke, MD, DSc
(Credit: University of Southern California)
As therapies become available for patients with Alzheimer disease (AD), studies show a growing concern from patients about wait times in getting diagnosis because of the limited accessible AD specialists and PET scanners. In a new study presented at the 2024 Alzheimer’s Association International Conference, July 28 to August 1, in Philadelphia, Pennsylvania, findings showed that using high-performance blood tests in primary care settings could potentially reduce wait times significantly and prevent eligible patients from falling outside the treatment window.1 Using this data, the researchers of the study plan to evaluate further scenarios and patient pathways as well as cost implications in ongoing research.
Conducted by lead author Soeren Mattke, MD, DSc, and colleagues, findings showed that if the patients had early-stage cognitive impairment and a blood test that indicated AD, 80% would be referred to an AD specialist. The specialist would then assess the patient and order biomarker testing, cerebrospinal fluid (CSF) testing (10%) and PET imaging (90%), for 90% of those who had confirmed early-stage cognitive impairment. Patients would then return to the AD specialist to discuss the results from the tests and talk about next steps in treatment. Investigators noted that if a blood test was used to confirm AD pathology, the patients would pass on the first visit to the specialist.
In a new iteration of NeuroVoices, Mattke, director of the University of Southern California (USC) Brain Health Observatory, had a conversation with NeurologyLive® to discuss how AD blood tests compare with traditional methods such as CSF analysis and PET scans in terms of diagnostic accuracy. He also talked about some of the challenges primary care doctors experience in incorporating AD blood tests into routine diagnostics. Moreover, Mattke, who also serves as a research professor of economics at USC, spoke about how the adoption of AD blood tests in primary care settings could potentially impact the treatment timeline for patients.
Soeren Mattke, MD, DSc: We have a big problem with AD because it is a common disease and almost all patients, particularly in the early stages, have not been formally diagnosed. There may be a recording of a memory complaint, but there's rarely a proper evaluation of where exactly they are in the disease continuum. There are very few cases where the pathology is confirmed with biomarkers. We have a backlog of prevalent cases that could be eligible for disease-modifying treatments, but we need to run them through a pretty complex diagnostic process. That process typically starts with primary care because it is a population-level disease, and we can't really start in specialty care.
Primary care doctors tend to perform a brief cognitive assessment like the Mini-Mental State Examination (MMSE) or Montreal Cognitive Assessment (MoCA) when patients come in with a memory complaint or if they suspect a memory issue. If the results are positive, they usually refer the patient to a dementia specialist, mostly neurologists. Very few primary care physicians conduct a full evaluation of the etiology of cognitive impairment themselves. That is still pretty rare; some do, but not many.
The problem with these referrals is that MMSE and MoCA have a relatively high false positive rate for mild cognitive impairment or mild dementia, with about 75% specificity. So, you are referring to about a quarter of patients who don't have any cognitive impairment—they just scored poorly on the test. Because this is such a large population, you are sending lots of people who don't need to see a neurologist, and since neurology is not a big specialty, this causes tremendous wait times. We argued back in 2017 when the first signals came along that lecanemab (Leqembi; Eisai) might work in AD, and we kept looking into this issue over the years as the treatments became more of a reality. Of course, nothing has changed. We still predict substantial wait times, and the anecdotal evidence from the early experience of lecanemab being available confirms this. If you look at the large memory centers, they now report 6, 8, and even 12 months wait lists.
Unfortunately, what we predicted is coming true. The question now is, what can we do about it? One thing that would help is better triage for AD in primary care, so that those who could be candidates for a disease-modifying treatment are referred with priority. A logical way, which we had already analyzed a few years ago, would be to use blood tests for AD. If you combine the information from an MMSE, which gives you reasonable evidence of cognitive impairment, with the information from a blood test, which gives you reasonable evidence of the presence of AD as the underlying etiology, you can prioritize patients for referrals much more efficiently. This would reduce the wait list and get the people who might be candidates for the treatment into appointments faster and then on treatment faster. You can probably handle most of the ones who do not have AD underlying etiology in primary care because we can't really offer them any causal treatment. We can offer symptomatic treatment, lifestyle changes, and adaptation to the disease. But for other etiologies, we don't really have causal treatments. That's the theory.
We tested this theory in a simulation model because the treatments haven't been around long enough to test it, and the blood tests are not really routinely available, especially not in primary care settings. We developed a Markov model to see how the insertion of a blood test into the early diagnostic process in primary care affects wait times down the road. We looked into two potential indications for that blood test. The first being a so-called rule-out test, where the test tells you “OK, if it's negative, it's pretty likely that you do not have AD underlying etiology, but it doesn't confirm that you have it. It just says, no, you're not having it. If it's positive, you still need to confirm.” Then we tested the other alternative, a so-called rule-in test or a confirmatory test that would replace the traditional confirmatory tests such as CSF analysis or PET scans for envoy.
We basically modeled these three assumptions and had to rely to a large degree on expert input because we don't really have these tests available in primary care. We don't have a lot of primary care doctors using them, so we had to use expert opinion as to how the test availability or lack thereof would influence decision-making. Based on their assumptions, we tested how the test will change decision-making, referral patterns, and also the speed to diagnosis in specialty settings.
One finding is that, if we look at the Medicare population because that is where the largest disease burden is going to be, the wait times begin relatively low at about a year but then keep growing as more and more patients enter the queue, and we can't really see them fast enough to reduce wait times over time. Over a 10-year horizon, they would actually grow to about 100 months on average if we just used MMSEs for referrals. In comparison, if you use the blood test, regardless of whether you do it for rule-in or rule-out, the wait time would grow only to about 14 months to diagnosis by the end of a decade. That's a pretty dramatic change in wait times.
What we have to keep in mind here is that we are talking about hard triage. We are saying, “well, you only get referred if both your MMSE is abnormal and your blood test is indicative of Alzheimer pathology.” Of course, in real life, you do not have such a hard triage. Some patients who are abnormal on the MMSE and negative on the blood test would still get referred to a neurologist for good reasons. Conversely, not all patients in whom both tests are indicative of early AD would get referred.
In the next step, we are planning to relax these assumptions and try to estimate how primary care doctors would add, not just referring the positive blood test, but some with positive blood tests and fewer with the negative blood test. That will give us a more realistic estimate of the effect on wait times for referrals.
The blood tests can truly revolutionize the diagnosis of AD and also the treatment monitoring. You can see whether the amyloid drugs remove the amyloid burden and stop at least the biological activity behind the disease. For that, we all need PET scans and CSF analysis today, and of course, they are complicated and not readily available everywhere. The blood tests can really revolutionize that diagnostic pathway.
It's also important that these blood tests are becoming precise enough, accurate enough to replace CSF and PET scanning, which would be an even greater step. This will take a little bit of time until specialists are comfortable enough with the blood test results that they say, “That's all I need to establish the diagnosis and potentially a treatment indication.” But of course, that would have an even greater impact than using this as a triage tool.
We don't know much yet about how these blood tests will play out in real-world settings. They are currently mostly used in specialist memory care, and there are several studies now underway that try to insert these blood tests into primary care and look into how they influence decisions. So, they don't just use expert opinion but real data on how primary care doctors act when they have the test or when they don't have the test. There are lots of exciting developments going on. I'm pretty sure next year at AAIC, we will have a lot more to say about the real-world impact of these tests.
Transcript edited for clarity. Click here for more coverage of AAIC 2024.
