With the advent of disease-modifying Alzheimer’s drugs, everyone in the world of Alzheimer’s diagnostics is bracing for the crush.
Based on their ability to slow cognitive decline, the Eisai/Biogen drug Leqembi won full approval from the FDA this month, and donanemab by Eli Lilly could be approved by year end. Clinicians now have to decide which patients are most suitable for treatment, which means better Alzheimer’s tests need to be developed.
“There’s going to be capacity issues because there’s just not enough neurologists in the world to treat this pool of patients with Alzheimer’s disease at the moment,” said Mark Stearman, a senior international product manager at Roche focusing on Alzheimer’s diagnostics.
Currently, diagnosing Alzheimer’s is slow. The standard confirmation test is an amyloid-PET scan, but by the time the disease is evident on the PET scan, the disease has already progressed, said Valerie Daggett, a bioengineering professor at the University of Washington. “A lot has to happen before you get there and you see those deposits,” she said.
Enter: blood tests. A glut of these tests — some used in clinical trials for Leqembi and donanemab — have received FDA breakthrough device designation, though none have been authorized yet. Newer tests and biomarkers are even trying to diagnose patients before they develop Alzheimer’s symptoms.
But as the market for these diagnostics opens up and researchers and clinicians increasingly lean upon these biomarkers to diagnose disease, it’s also causing researchers and neurologists to question what exactly Alzheimer’s is. Last week, the National Institute on Aging and the Alzheimer’s Association presented new guidelines redefining Alzheimer’s, partially because of the advent of blood biomarkers.
“This is, I think, not the 1 million [dollar], but the 10 million [dollar] question for Alzheimer’s disease: What is Alzheimer’s?” said Tamas Fülöp, a professor of medicine and geriatrics at the University of Sherbrooke in Canada. “Is this only clinical, is this only pathological? Is this both?”
Joseph Quinn, a professor of neurology at Oregon Health & Science University, agreed. “The field and the world still has to come to grips with the idea that the pathology in the brain exists even before there are symptoms and exactly how we’re going to describe people who have biomarker evidence of Alzheimer pathology but no symptoms at all.”
How it works
These blood tests are based on the idea that Alzheimer’s is caused by a protein fragment called beta-amyloid, which eventually clusters into clumps called plaques. Thus, the amount of amyloid people have floating freely in their cerebrospinal fluid (CSF) or blood should decrease if they have Alzheimer’s and the beta-amyloid is clumping into plaques. Neurofibrillary tangles are also thought to have a role in Alzheimer’s and other neurodegenerative diseases. These filaments form around neurons when tau proteins become altered by a process called phosphorylation and end up aggregating together. Thus, finding elevated amounts of phosphorylated tau, or ptau, in blood or CSF is also a way to measure Alzheimer’s.
Each person has a bit more than a tenth of a liter of cerebrospinal fluid, cushioning and hydrating their brain and spinal cord at any given moment. As one of its functions, the fluid helps carry waste out of the brain — including the bits of proteins used as biomarkers for neurodegenerative diseases like Alzheimer’s. Practitioners can sample a patient’s CSF through a lumbar puncture, which isn’t the ideal or easiest procedure for routine testing. However, the biomarkers in the CSF are clear and strong enough that this procedure can be used to confirm an Alzheimer’s diagnosis.
CSF releases its waste cargo into the bloodstream for further breakdown and elimination, which means those same biomarkers can be detected in blood. But they get diluted in the bloodstream, since a person has about five liters of blood, about 40 times as much volume as a person’s CSF.
One reason these blood tests can’t be run at individual doctor’s offices and are currently evaluated at either a company’s centralized facility — in the case of St. Louis-based C2N Diagnostics, for example — or centralized labs with trained staff on the company’s instruments, as in the case of Roche or biomarker analysis company Quanterix, is that these dilute bits of amyloid and ptau are hard to detect.
“Whereas in the CSF, it’s a nice clean fluid that you can measure, in the blood it’s a bit of a minestrone soup and we are measuring these compounds at very low concentrations with lots and lots of background noise because there’s so many other proteins in the blood,” said Stearman at Roche. “So it is challenging to take these samples into the blood.”
The need
Because Leqembi and the Lilly antibody donanemab both aim to get rid of amyloid, patients need to confirm that they have amyloid buildup before they’re eligible for the treatment. That means they will need both a scan to get diagnosed, and to monitor if the treatment is working, which is more than the current number of scans the Centers for Medicare and Medicaid Services covers: one. But just last week, CMS proposed eliminating one hurdle to PET access — loosening its previous policy to allow patients to get multiple scans, and get them outside of the context of a clinical trial.
The trouble is there are only about 2,000 PET centers in the United States, and they’re heavily booked, said Masoud Toloue, CEO of Quanterix, which makes an Alzheimer’s blood test.
Using simpler screenings to decrease the barriers to getting an Alzheimer’s diagnosis could make it easier to get people treatment earlier in the course of the disease, said Maria Glymour, chair of the epidemiology department at Boston University, in an email to STAT. She warned this would pay off only if healthy people are not incorrectly diagnosed with the disease, and if there are Alzheimer’s treatments with substantial, long-term benefits.
“As of today, we have treatments with clear evidence of a small, short-term benefit. We don’t have good evidence on whether the short-term benefit stays the same, grows, or disappears over the longer run,” she said.
Regardless of whether new drugs will truly be highly sought after — Quinn at Oregon Health and Science University says he’s seen far less enthusiasm from his patients than he anticipated — diagnostics companies are preparing for the floodgates to open.
Many people have started to talk to companies like Roche about how they can implement biomarker testing at the local level, said Stearman. Roche earlier this year announced a partnership with Lilly to develop an assay that looks at Alzheimer’s blood biomarker ptau181 and the most common genetic risk factor for Alzheimer’s, APOE4. Lilly is mostly supplying the finances, said Stearman, while Roche is making sure “that we can identify as many patients as possible.” Lilly has also partnered with Quanterix on a test to measure ptau217, another Alzheimer’s biomarker.
In anticipation of more people needing test results on their way to getting a Leqembi prescription, Quanterix launched a patient-facing version of its test earlier this month. The rebranded test is now called LucentAD and has a patient portal, blood collection logistics, and other infrastructure needed to sell the test directly to patients and physicians instead of to researchers. However, those looking to get the test will have to pay out-of-pocket for now. C2N quoted a list price of $1,250 for its PrecivityAD test, and noted that financial assistance programs are available. Testing companies are working on getting CMS to cover their blood-based tests.
As blood tests edge out of clinical trials, Sebastian Palmqvist at the University of Lund in Sweden has started investigating how these blood tests will perform under real-world conditions. Patients will not be coming exclusively from specialists at memory clinics, where many research studies currently obtain samples and patients. Instead, people will be asking their primary care providers whether their memory problems are Alzheimer’s. Samples won’t be processed at testing labs all in one batch, as research samples are, but will trickle in, adding day-to-day variation to the sensitive results. Encouragingly, Palmqvist didn’t see significant differences in the results from samples he shipped at different times.
Palmqvist’s research, presented at AAIC last week, tested C2N’s PrecivityAD2 Alzheimer’s blood test against Swedish primary care physicians’ diagnoses. The team found that when verified against CSF testing, the physicians diagnosed Alzheimer’s correctly only 55% of the time, compared to more than 85% for the blood test. Palmqvist’s research team also assessed how sure the clinicians were that Alzheimer’s was the cause of their patient’s memory problems, and found that they were less than 50% sure of their diagnoses, which caused more than 50% of people who had Alzheimers to not receive treatment and around 30% of people who did not have Alzheimer’s to incorrectly receive treatment.
Companies that spoke with STAT emphasized that blood tests are triage tests for now, merely indicating which patients should go on to get a confirmatory PET scan or CSF test, and shouldn’t be used for diagnosis on their own. Palmqvist proposed at the meeting that test developers could create cutoffs wherein patients’ results would fall into yes, no, and maybe ranges, and only the results in the gray zone would need confirmatory testing.
However, the enormous financial and risk burden from current Alzheimer’s drugs necessitates confirmatory tests, at least in the short term, said Palmqvist. “I know that at least in my clinic, which is the largest one in Sweden, we will consider the treatment to be so expensive and potentially also harmful for the individual that we will, at least in the beginning, always require confirmation with CSF or PET if it comes to these expensive treatments,” he said.
Easy access to home tests for Alzheimer’s worries Quinn. As a clinician at OHSU, he’s seen marketing materials that suggest that anyone can get their hands on these tests and get a yes or no answer. It seems simple to anyone who wants to know, especially as tests get better at diagnosing people early. But what about people who aren’t symptomatic yet but test positive for amyloid or tau?
“That’s where the availability of home tests for Alzheimer’s disease become challenging because you can have somebody who may have biomarker evidence of Alzheimer pathology, no symptoms at all, maybe they’re not going to be symptomatic for a decade or more. And if they get the information that they have Alzheimer’s disease, it can lead to suboptimal life decisions,” said Quinn.
It’s a weighty diagnosis, which is why clinicians, especially primary care physicians, can take a long time to put a name to the disease in the hopes that a few memory problems are just a sign of normal aging. But when treatment predicates on having the biological hallmarks of Alzheimer’s, but someone with the pathology doesn’t have signs of cognitive impairment yet, do they have Alzheimer’s?
What is Alzheimer’s
At the Alzheimer’s Association International Conference last week, the National Institute on Aging and the Alzheimer’s Association presented new guidelines proposing that Alzheimer’s should be defined biologically, not based on clinical symptoms. The working group said the updates responded to many changes in the Alzheimer’s landscape in the last five years, mainly advances in blood-based biomarkers as well as the arrival of approved treatments for the disease.
“Historically, the term Alzheimer’s disease has been reserved for people who have an overt dementia, things that are impairments that are advanced enough to interfere with their everyday activities,” Quinn said. “In more recent times, the term has been extended to cover people who have milder impairments but do have biomarker evidence of Alzheimer pathology, so the term that’s often used to describe those people is ‘mild cognitive impairment due to Alzheimer pathology.’”
Palmqvist agreed with the reasoning behind the new guidelines, noting that we don’t require a person to be in the symptomatic phase of the disease to “have” a disease.
“We, for example, call it prostate cancer, even though the majority of men with prostate cancer will never have any symptoms from the cancer,” he said. “I think we need to normalize the term Alzheimer’s disease, but still use it” — even though this new definition will increase the prevalence of Alzheimer’s.
Fülöp at the University of Sherbrooke doesn’t ascribe to the amyloid hypothesis that’s become entrenched in the field. He sees beta-amyloid as not the cause, but the consequence.
“Beta amyloid, neurofibrillary tangle surrogates like first phosphotau181, 217, 231, whatever is the number…everybody’s screaming, ‘This is the most important, the most accurate measure of the Alzheimer disease!’” he said. But because Alzheimer’s is a late manifestation of the disease, which is something the new NIA-AA guidelines concur with, Fülöp thought these biomarkers might be useful if treated as hallmarks only of late-stage disease, and if other biomarkers were discovered to define more preliminary stages.
A new study published in Science Translational Medicine attempted to do just that: researchers from Johns Hopkins and the NIA found 32 proteins linked to early-onset dementia that might be used to detect Alzheimer’s early. Daggett’s lab at the University of Washington found another early indicator of Alzheimer’s, toxic oligomers of beta-amyloid, that could also be used as an early treatment target. Daggett’s company AltPep has received an FDA breakthrough device designation as they work on commercializing the test.
Lilly is using a ptau217 blood test to identify patients who have amyloid pathology but are still not showing Alzheimer’s symptoms for its TRAILBLAZER-ALZ3 trial of donanemab. The expectation is that the drug might be more effective on people who are in earlier stages of the disease, and these patients could potentially have a lower risk of ARIA, said Akash Tewari, an analyst at Jefferies, in a research note. The trial started in 2021 and is expected to have results in 2027.
Other areas researchers are working on include further simplifying blood tests to fingerprick tests that could be spotted onto a card and mailed in, instead of getting a blood draw. Hanna Huber, a postdoc in the neurochemistry department at the University of Gothenburg in Sweden, presented a study at AAIC that showed that fingerprick tests work just as well as venous blood draws.
Right now, blood tests are “rule-out” tests, tests that make sure that the wrong people aren’t getting further testing and treatment. But the holy grail would be a rule-in test, said experts. Toloue said Quanterix is in “advanced clinical trials” for such a blood test, which includes multiple biomarkers and would be robust enough to not need confirmation from PET or CSF.
Even within treatments, there are diagnostic opportunities, said Stearman at Roche. That includes figuring out who is more prone to ARIA, the brain-swelling and brain-bleeding that can be a serious side effect of amyloid-targeting treatments.
Even though the advent of treatments seems to signify that the field has figured out Alzheimer’s disease, the continual discovery of new fluid biomarkers has only shown that researchers have more to discover in the Alzheimer’s landscape.
“This is so like a puzzle. We have many small pieces we tried to put together, but the whole picture is still not there,” Fülöp said, frustrated by the disjointed nature of research that fixates on minutiae instead of the bigger picture. “Our aim is to find something for the patient! So it’s not a theoretical debate. This is very, very important [when] I have a patient in front of me and he’s asking me, ‘What can I do?’”