23 Apr 2025

In memory clinics and in research cohorts, immunoassays for plasma markers can now distinguish people who have Alzheimer’s disease pathology with remarkable accuracy (see Part 7 of this series). How about in the general population? Could these tests catch people in the early stages of AD, before they turn up at their doctor with memory concerns?

Scientists are starting to test this idea in real-word situations. At AD/PD earlier this month in Vienna, Michael Schöll, University of Gothenburg, Sweden, gave an update on the REAL AD study in Västra Götaland in the southwest of the country, home to about 750,000 people. “Our cohort is unique because we focused on implementation and integration into the existing healthcare infrastructure,” he said.

Västra Götaland boasts advantages for this type of study. Electronic healthcare records from the Swedish universal health care system are good, and society is “highly digitized across all ages,” said Schöll, i.e., most everyone is comfortable using computers, smart phones, and apps. In fact, in Sweden people must bank with a phone app, and their BankID serves as official identification. REAL AD capitalized on this to recruit, monitor, and assess volunteers (see Jan 2024 news).

The study started a year ago. Since then, more than 7,000 people who saw the website or pamphlets at their doctors’ offices signed up. Of these, just under 5,000 completed initial cognitive testing called the Quick Dementia Rating System. Thirty-five hundred and fifty-one decided to take part in longitudinal cognitive screening, which involved their choice of three cognitive tests three times over three months on a phone app called neotiv (Nov 2020 news), or six test sessions over three months on a personal computer app called Cognitron. Almost 90 percent opted for the phone app. Almost 2,000 answered optional questionnaires about sleep patterns, drug use disorders, alcohol use, diet, physical activity, depression, and anxiety. They were invited to give a blood sample at any of 111 sampling locations throughout the region, which then shipped the samples to a central lab in Gothenburg for analysis. Tests were done on Roche Elecsys or Beckman-Coulter machines, whichever was available as samples came in, Schöll said. So far, his team have 2,839 sample in the freezer for analysis.

What did they find? Schöll shared some hot-off-the-press data. “We’ve only had it a few days, so it is very preliminary,” he told the audience in Vienna. Still, some patterns may be emerging. The neotiv https://neotiv.com/en/studies cognitive test appears to correlate with age, with older people scoring lower. Men performed worse than women across all ages; though the difference was slight, it was consistent, said Schöll.

Intriguingly, plasma p-tau217 rose with age, and more so among people who tested positive for amyloid. This was slightly different than Sebastian Palmqvist, Lund University, reported for primary care settings. There too, the marker inched up with age in people who were amyloid-negative, but the opposite was true among the amyloid-positives (see Part 7 of this series). In REAL AD, baseline numbers were slightly higher in people who carried an ApoE4 allele.

Using two cutoffs, Schöll found that people in the low p-tau217 category tested highest on the neotiv cognitive test, those with intermediate p-tau217 did slightly worse, and those with the most plasma p-tau217 had the lowest cognitive scores. Using a single cutoff, people deemed cognitively unimpaired still tested significantly lower on neotiv if they were p-tau217 positive. The upshot? “Even in this ‘dirtiest’ of cohorts, we find an association between self-reported memory impairment and plasma p-tau217,” said Schöll.

This study was well-received among an AD/PD crowd hungry for real-world data. “This is truly exciting to see such a digitized enrollment,” Palmqvist said. “You are at the forefront of this topic.” One attendee, calling the work impressive, asked if these participants can enroll in a memory clinic or a trial. Alas, since these plasma tests are not yet approved and none of the cutoffs or thresholds have been clinically validated, REAL AD is not sharing data with the volunteers, Schöll replied. “Ethically, it has been a difficult issue because people are asking for their results,” he said. Once the tests are cleared for use, he plans to go back and inform them.

More broadly, digital measures are creeping into the AD field. In his talk, Oskar Hansson, Lund University, Sweden, sounded excited about the idea of combining blood markers with a digital cognitive assessment. He has developed such a test, called BioCog. The idea is that an elderly person with a cognitive complaint would take this test and, if it suggested impairment, get a blood test. If not, they’d be deemed cognitively unimpaired without further testing. BioCog runs on a tablet and is self-administered, explained Pontus Tideman, a graduate student at Lund. It takes five to 15 minutes, uses a word list to test immediate and delayed recall, and incorporates delayed recognition and cognitive processing speed tasks. It also asks what day, date, month, and year it is. This helps weed out people who have more severe memory issues, explained Tideman.

Blood and Digits. Combining a digital cognitive test called BioCog with C2N’s APS2 plasma test improved diagnostic accuracy in a primary care setting. [Courtesy of Oskar Hansson, Lund University, Sweden.]

Hansson has evaluated this approach in the clinic. First, he set a cutoff for cognitive impairment on BioCog using a discovery cohort, then evaluated it in a primary care setting. BioCog alone beat a primary care physician in diagnosing AD, getting it right 85 percent of the time versus 73 percent for the doctor, for positive and negative predictive values of 87 and 83 percent, respectively. Adding on the plasma APS2 test from C2N diagnostic bumped up accuracy to 90 percent and 95 percent for the one- and two-cutoff approaches, respectively. APS2 alone, a combination of plasma %p-tau217 and Aβ42/40, got it right 80 percent of the time (image above). “I think this is promising,” said Hansson. “Of course, it is only one study in primary care, but I would love to see other studies like this in the future.”

Across the field, primary care and other real-world studies are springing up to figure out ways to bring immunoassays and digital tests into routine care. BioFinder Preclinical also addresses this in Sweden, and ADNI 4 in the U.S. has adopted remote digital and plasma marker screening to find people for more thorough follow-up who might have early AD. Hansson was optimistic, but cautioned that when moving into very early AD, ratios might be the better choice for plasma markers. “There is less pathology in the brain and changes in biomarkers are more subtle, so correcting for variation due to comorbidities such as chronic kidney disease becomes more important,” he said. Ratios, such as p-tau217/tau217, help in this respect.—Tom Fagan

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

1. Looking Good: Immunoassays for Blood Markers 23 Apr 2025
2. AD Blood Tests Are Here. Now, Let's Grapple With How to Use Them 25 Jan 2024
3. Learning Troubles Spied by Smartphone Track with Biomarkers 20 Nov 2020

External Citations

1. https://neotiv.com/en/studies
2. BioFinder Preclinical
3. ADNI 4

Further Reading

No Available Further Reading