An experimental blood test was as accurate at detecting Alzheimer's disease in patients as much more costly and invasive procedures, according to a new JAMA study—findings that could soon make diagnosing the disease easier, more affordable, and more available.

New case studies: Keep Alzheimer's patients safe at home and in the community

Current Alzheimer's screenings can be inaccurate, costly

According to the New York Times, to diagnose Alzheimer's, doctors primarily rely on clinical assessments of memory and cognitive awareness, as well as conversations with patients' loved ones. However, these diagnoses are frequently inaccurate, because it can be difficult to differentiate symptoms of Alzheimer's from other types of dementia or from physical conditions that can impair cognitive capacity, the Times reports.

Moreover, other clinical assessments for the disease, such as PET scans and spinal taps, are costly, difficult to access, and imprecise when it comes diagnosing Alzheimer's. Those tests focus on detecting elevated levels of amyloid protein, which forms plaques in the brains of Alzheimer's patients. However, the detecting high levels of amyloid alone can't serve as a diagnosis of the disease, because some people can have high amyloid levels without having Alzheimer's.

Researchers evaluate a blood test for Alzheimer's

But for the study published Tuesday in JAMA, researchers focused on a blood test aimed at detecting Alzheimer's that measures elevated levels of a different protein, called tau, which is found in tangles throughout the brains of patients with the disease. One form of tau, p-tau217, is a reliable indicator of Alzheimer's and, according to a study recently published in the Journal of Experimental Medicine, p-tau217 correlated more closely to buildups of amyloid than other forms of tau.

For the JAMA study, Oskar Hansson, a professor of clinical memory research at Lund University in Sweden, and colleagues researched the efficacy of the blood test measuring elevated levels of p-tau217 in 1,402 individuals who were already enrolled in dementia studies in Arizona, Colombia, and Sweden. The participants included individuals without any cognitive impairment, those with mild impairment, those with Alzheimer's, and those with other neurological diseases.

The blood test evaluated in the study was developed by Eli Lilly, which provided the researchers with the materials and a few employees to perform the assays. Eli Lilly, although allowed to review the manuscript, was not permitted to veto any findings. The study's funding came primarily from government entities and foundations in Sweden and the United States, the Times reports.

Key findings

According to the study, which was shared this week at the Alzheimer's Association International Conference, the blood test identified patients with Alzheimer's with 89% to 98% accuracy.

Specifically, in the part of the study focused on individuals in Sweden, researchers found that the test was 96% accurate in distinguishing whether people with dementia had Alzheimer's or a different type of neurodegenerative disorder, Hansson said. According to the study, that level of performance was comparable to PET scans and spinal taps in diagnosing the disease, and it outperformed MRI scans and blood tests focused on amyloid, a different type of tau, and a third neurological biomarker for the disease called neurofilament light chain.

Overall, according to Hansson, the blood test found that people with Alzheimer's had seven times as much p-tau217 as people who either did not have dementia or who had different neurological diseases, such as Parkinson's disease—making the test very "specific for Alzheimer's disease."

In the Arizona portion of the study, researchers compared findings from brain autopsies on donors with test results from blood the donors gave before death. The researchers found that the blood test was 98% as accurate at diagnosing Alzheimer's as autopsies of people found likely to have the disease because of amyloid plaques and extensive tau tangles in their brains. The test also was 89% as accurate as autopsies of people considered "moderately likely" to have Alzheimer's because their brains had amyloid plaques but fewer tau tangles, the Times reports.

In addition, for the portion of the study conducted in Colombia, researchers looked at a family with a rare gene that nearly guarantees a person will develop Alzheimer's and found that levels of p-tau217 began rising in those individuals "around 20 years before symptoms" of the disease started to occur, according to Hansson.

Meanwhile, in a separate study from the University of California-San Francisco (UCSF) that was also presented at the conference, researchers found that the same type of blood test used in the JAMA study was able to distinguish Alzheimer's patients from those with a different neurological disease, frontotemporal dementia, with 96% accuracy in a study involving 617 people.

'It's a new world'

The studies' results suggest that a simple blood test could replace current methods of testing for Alzheimer's that can be invasive, costly, or imprecise, the Associated Press reports. According to the Times,  experts also said the test may accelerate screening for clinical trial participants—a process that currently takes several years and requires millions of dollars to complete because of reliance on more expensive testing methods—which in turn could speed up the search for and development of new therapies for Alzheimer's disease.

"This test really opens up the possibility of being able to use a blood test in the clinic to diagnose someone more definitely with Alzheimer's," Maria Carrillo, chief science officer at the Alzheimer's Association, said. "Amazing, isn't it? I mean, really, five years ago, I would have told you it was science fiction."

"This blood test very, very accurately predicts who's got Alzheimer's disease in their brain, including people who seem to be normal," Michael Weiner, an Alzheimer's disease researcher at UCSF, said. "It's not a cure, it's not a treatment, but you can't treat the disease without being able to diagnose it. And accurate, low-cost diagnosis is really exciting, so it's a breakthrough."

Eliezer Masliah, chief of neuroscience at the U.S. National Institute on Aging, said the results are "very encouraging" and the new testing "appears to be even more sensitive and more reliable" than other methods, but it still needs to be tested in larger, more diverse populations.

Researchers said the test would need to be refined, standardized, and approved by federal regulators, but estimated it could be available for clinical use in two to three years—and, according to Rudolph Tanzi, a professor of neurology at Massachusetts General Hospital and Harvard Medical School who was not involved in the study, it "has the promise to make early detection of the disease possible, before we have symptoms."

"Within a few years, it's very possible that there will be certified laboratory tests for these proteins and others, and maybe tests will be developed for Parkinson's disease and so forth," Weiner said. "It's a new world" (Marchione, Associated Press, 7/28; Belluck, New York Times, 7/28; Owens, "Vitals," Axios, 7/29).