In normal function, tau proteins (red) stabilize the microtubules (blue) that support neuronal axons. In Alzheimer's disease, tau gains phosphate groups (yellow) and can instead form pathological tangles.Credit: Juan Gaertner/ Shutterstock
The treatment of Alzheimer’s disease (AD) is in a new era with the introduction of drugs that impact the underlying biology. These therapies need to be administered early in the disease course to have the most benefits, so identifying people in these early stages has taken on new urgency.
Biomarker-based diagnostics have also markedly improved in recent years. The current gold standards — PET scans of brain amyloid and analysis of proteins in cerebrospinal fluid (CSF) obtained from lumbar punctures — can identify people with AD pathology with high accuracy. But these methods are invasive, expensive, and not widely available, particularly in economically disadvantaged regions.
Clinicians need more convenient tools for identifying patients if the promise of new treatments is to be fully realized. “We also have to make sure the public knows that Alzheimer’s is no longer a diagnosis with no treatment,” says Andrew Liu, a neurologist at Duke University in Durham, North Carolina. “And that not only can we diagnose them, but we also have disease-modifying therapies now.”
Telltale blood
A solution exists in the shape of blood-based biomarkers (BBMs). Pathological processes occurring in the brain produce proteins that leak into the bloodstream, and so are available from a simple blood draw.
For AD, most people think of amyloid-Beta, but there is another protein called p-Tau (phosphorylated tau). The tiny amounts of p-Tau that leak into blood were extremely difficult to detect using previous tools, but advances in immunoassay and mass spectrometry technologies have now delivered tests that are able to measure these proteins.
Of the two, p-Tau has emerged as the more specific and robust biomarker — at least in blood samples. Although the ratio of the two isoforms of amyloid-Beta (aB42/40) changes significantly in CSF, the difference in blood is not as large. “In plasma, amyloid levels decrease only about 10-15% between controls and AD, probably due to peripheral sources that also produce amyloid,” says Charlotte Teunissen, a neuroscientist at Amsterdam University Medical Center, Netherlands. “P-Tau can increase three- to seven-fold1, so it’s a very robust biomarker.”
P-Tau is also better than aB42/40 at distinguishing between AD and other neurodegenerative diseases. “P-Tau specifically increases in AD and not in other tauopathies,” Teunissen says. What’s more, plasma p-Tau is tightly correlated with multiple important outcomes, including amyloid burden, tangle burden, cognitive decline, and brain atrophy, Teunissen adds. And, importantly, it appears in blood very early in the disease process2,3.
There are several isoforms of p-Tau. However, multiple studies have demonstrated that p-Tau 217 is currently the best predictor, and can diagnose AD with an equivalent performance to measuring it in CSF4.
New science
In 2018, the National Institute of Aging and the Alzheimer’s Association (AA) published recommendations for the diagnosis and staging of AD, which acknowledged the central role of amyloid and tau pathology in disease progression. The AA released a draft updated document for public consultation and feedback last year, revising these diagnostic criteria in light of the latest science and test capabilities. “The strong performance of plasma p-Tau 217 tests is one reason the criteria have been revised,” says Teunissen, who is part of the group tasked with updating the criteria. “Blood biomarkers play a prominent role now.”
Lucent Diagnostics, part of Quanterix, recently launched a digital p-Tau 217 immunoassay based on Quanterix’s Simoa (single molecule array), coupled with tau-specific antibodies. Conventional immunoassays must detect target molecules diluted in solution, but the Simoa platform uses microwells to isolate individual immunocomplexes (consisting of an antibody-bound target protein and fluorescent reporter enzyme), boosting sensitivity at least 100-fold over conventional analogue immunoassays5.
Plasma p-Tau 217 can detect amyloid pathology with an accuracy exceeding 90%, which meets the AA’s revised criteria for biomarker tests for diagnostic purposes. Laboratory-developed tests analysing BBMs, including p-Tau, are commercially available in the United States. Lucent’s test, LucentAD, aids in the diagnostic process by evaluating whether someone with changes in cognition has amyloid pathology consistent with AD.
BBMs are already being used to screen potential participants for clinical trials. These non-invasive and cost-effective blood tests drastically reduce the costs of recruiting participants in the earliest stages of AD, while the high accuracy of plasma p-Tau 217 reduces the need for many PET or CSF tests in clinical trials.
BBMs may also be useful in monitoring disease progression, or even treatment response. A phase III clinical trial6 of one monoclonal antibody, approved by the FDA in 2023 for AD, found that treatment reduced levels of p-Tau 181. “That was the convincing evidence that disease modification really occurred, because p-Tau is the one that correlates most with symptoms,” says Liu. In another phase III trial7 for a different antibody, researchers also saw decreases in p-Tau 217.
Remaining reservations
A few questions remain. Some studies have reported differences in plasma p-Tau levels in different ethnicities1, but whether this relates to socioeconomic differences, comorbid medical conditions, or something else, is unclear.
As with any BBM, biological factors such as BMI, kidney function, cardiovascular disease or peripheral neuropathies, can affect p-Tau levels. “I had a patient with positive plasma biomarkers and he was terribly afraid that meant he had AD,” says Liu. “But he had chronic kidney disease, hypertension and diabetes, which all modify plasma results.” Physicians need to be aware of the potential for these factors to influence blood test results.
Researchers also need to confirm that BBMs are equivalent to the more invasive methods in the messy world of clinical practice, with diverse patients, measurement errors, and possible confounding factors. “We’re doing prospective studies to build evidence and gain confidence,” Teunissen says.
Once the details are ironed out, BBMs look set to revolutionize AD diagnosis and monitoring. For one thing, blood tests are easy to administer; small community hospitals typically don’t have PET scanners. “We’re really addressing the equity of diagnosing Alzheimer’s disease accurately,” Liu says. “That’s why blood-based biomarkers are such a game changer.”
P-Tau and amyloid are not the only signs of neurodegeneration in blood. Others, including NfL (neurofilament light chain protein, indicating damaged axons), and GFAP (glial fibrillary acidic protein, indicating neuroinflammation), are elevated in many neurodegenerative diseases. They may not distinguish between conditions well, but combinations of biomarkers may be useful for making broader clinical diagnoses, or for prognosis. “GFAP is prognostic for all-cause dementia,” says Teunissen. “We need combinations of markers to make a better clinical diagnosis for different dementia types.”
Researchers are following the same paths for other conditions. “Ultimately, I expect we’ll have blood biomarkers for every neurological disease,” Teunissen predicts.