When the U.S. government approved the Alzheimer’s disease drug aducanumab last month despite shaky evidence of clinical benefits, Suzanne Schindler saw an immediate consequence: “We’re going to have to do a lot more biomarker testing.” Schindler, a neurologist at the Washington University School of Medicine in St. Louis, expects many patients with memory problems will want to know whether they are eligible for the drug, the first meant to address the underlying disease process and slow cognitive decline. But a definitive diagnosis is expensive and time-consuming, involving a hunt for indicators of disease, or biomarkers, in patients’ spinal fluid or in positron emission tomography (PET) scans. “We simply don’t have the capacity” to do the tests on everyone potentially eligible for aducanumab, Schindler says.
An innovation could help manage the deluge: simple blood tests for molecules that may indicate signs of disease in the brain. “A lot of data now [suggest] that what we find in the blood actually can reflect what’s going on in the brain,” says Nancy Ip, a neuroscientist at the Hong Kong University of Science and Technology. Her team developed a test to identify Alzheimer’s and determine a patient’s stage of disease from levels of 19 blood proteins.
One blood test is already commercially available in the United States, though some researchers doubt it can meet demand. Companies are now working to make other tests fit for widespread clinical use—and doctors are scrambling to decide when and how to use them. Before the FDA decision, many physicians “had a somewhat leisurely approach” to determining how Alzheimer’s blood tests might fit into their clinical practice, says Douglas Galasko, a neurologist at the University of California (UC), San Diego. “And now life is much more complicated.”
Aducanumab, approved to treat early-stage disease, reduces buildup in the brain of beta amyloid, a sticky protein thought to drive neuronal damage in Alzheimer’s. But its evidence of clinical benefit is thin, the drug requires monthly infusions with a list price of $56,000 per year, and it can cause swelling and bleeding in the brain. Deciding to prescribe it “is a really big deal,” Schindler says. “You want to be absolutely certain [patients] have Alzheimer’s disease.”
Either a PET scan that detects beta amyloid in the brain or a spinal fluid test that measures amyloid and another Alzheimer’sassociated protein, tau, can confirm that diagnosis. However, PET scans can cost thousands of dollars, and many patients dread a lumbar puncture to collect spinal fluid. And if every older person with cognitive impairment required evaluation by a dementia specialist and a PET scan or spinal fluid test, they would quickly overwhelm the medical system, says Soeren Mattke, a health services researcher at the University of Southern California’s Center for Economic and Social Research.
In a study published last year, he and colleagues predicted that the first disease-modifying Alzheimer’s drug available in the United States would lead to wait times of up to 45 months for a specialist evaluation and confirmatory tests. But blood tests that could screen out patients who don’t have Alzheimer’s could reduce wait times to about 10 months, the researchers found.
Reliably capturing the state of the brain from proteins that slip into blood has been a challenge. Blood contains so many different proteins that “you’re looking in what is essentially very dirty pond water,” says Ashvini Keshavan, a neurologist at University College London. Beta amyloid hides from detection by clinging to surfaces or to other proteins—and it can quickly degrade in a blood sample, adds Henrik Zetterberg, a neurochemist at the University of Gothenburg.
One type of blood test uses an instrument called a mass spectrometer to measure the ratio of two types of beta amyloid in blood. Decreased levels in blood of certain forms of the protein indicate it is instead accumulating in the brain as plaques. In 2019, researchers reported that a combination of this ratio, a person’s age, and a measure of genetic Alzheimer’s risk agreed with results of a PET scan in 94% of cases.
That test, marketed by C2N Diagnostics, has been available to U.S. clinicians since October 2020 under a Centers for Medicare & Medicaid Services (CMS) certification process for laboratories that test patient samples. The test costs $1250; a financial assistance program can bring that to $25 to $400, says C2N CEO Joel Braunstein. The test must be done in the company’s lab, which can run 250,000 to 300,000 tests per year, Braunstein says.
Other types of blood tests, so-called immunoassays, use antibodies to sandwich and isolate a protein of interest. Some might run on equipment common in clinical labs. Many recent studies of these tests have focused not on beta amyloid, but on certain forms of tau. Beta-amyloid plaques can appear decades before patients develop dementia—if they ever do. But beta-amyloid buildup is thought to somehow prompt tau to form toxic tangles that damage neurons, explains William Jagust, a neuroscientist at UC Berkeley. As a result, he says, “It may well be that some of these tau biomarkers are better predictors” of symptom onset.
In May, Lund University neurologist Oskar Hansson and colleagues reported in Nature Medicine that combining a tau blood test with a person’s Alzheimer’s genetic status and two brief cognitive tests could predict with about 90% accuracy which people with mild cognitive impairment would develop Alzheimer’s dementia within 4 years. “That was really groundbreaking,” Keshavan says.
Several companies have developed blood tests sensitive to tau, including Roche and Eli Lilly and Company. A large clinical trial of Eli Lilly’s antiamyloid antibody donanemab will use a tau blood test to help select patients. But to be used as clinical decision tools, the tests need to show consistent results under variable conditions and in more diverse patient populations, Hansson says. Most studies have used blood samples from patients at specialized memory clinics. Many of these groups are disproportionately white or of high socioeconomic status; and they generally don’t have other conditions, such as heart and kidney failure, which could influence levels of certain biomarkers, he says.
Some researchers caution that blood tests aren’t ready to replace a PET scan or spinal fluid test. And confirming Alzheimer’s will still require careful clinical examination, Zetterberg says, because many factors can cause memory problems in aging patients, including depression and sleep disruptions. Prescribing the new drug to anyone with a positive blood test is “the nightmare scenario of starting to use the biomarkers in a careless way,” he says.
But blood tests could act as an initial screen. A test in primary care offices would offer peace of mind to those who test negative—and could reduce the number of people seeking a PET scan or spinal fluid test such that CMS might cover these costs, says Sid O’Bryant, a clinical neuropsychologist at the University of North Texas Health Science Center. “Without a blood test, access to the drug is going to be so incredibly limited,” he says. He expects such tests to be available within the next 2 years. “The data and the science are so close.”