Blood Biomarker Disclosure Doubles Alzheimer Diagnostic Certainty in Randomized Trial
Sharing p-tau217 and NfL results early boosted high-confidence Alzheimer diagnoses from 4.6% to 50% without worsening patient anxiety.
Summary
When people have memory complaints, doctors often cannot be sure what is causing them. This Spanish randomized trial tested whether sharing results from an Alzheimer's blood test (p-tau217, plus neurofilament light chain) earlier would help. Among 220 memory clinic patients, those whose results were shared at 3 months rather than 9 months were far more likely to receive a diagnosis their neurologist was at least 90% sure of (50.0% vs 4.6%). They were also more likely to start Alzheimer's symptom medication and be sent back to primary care, and less likely to need repeat cognitive testing. Importantly, earlier disclosure did not raise anxiety, depression, stress, or lower quality of life. The findings support using blood tests as part of routine memory clinic workups.
Detailed Summary
Working out why someone is having memory problems is difficult. Clinical evaluation alone is imperfect compared with neuropathology, and the reference tests, spinal fluid analysis and amyloid PET, are costly, invasive, or hard to scale. Plasma phosphorylated tau at threonine 217 (p-tau217) is the most accurate blood marker of Alzheimer disease pathology, but accuracy does not prove clinical usefulness. Until now, evidence on diagnostic certainty, management, and emotional impact came largely from studies without randomized controls.
The PLASMAR trial enrolled consecutive new outpatients at the Hospital del Mar memory clinic in Barcelona between February and October 2024. Of 384 people screened, 265 were eligible, 229 consented, and 220 were randomized (median age 73; 55.9% women). Baseline diagnoses were subjective cognitive decline (45.5%), mild cognitive impairment (30.9%), and mild dementia (23.6%). Participants were assigned 1:1 by minimization to have plasma p-tau217 and neurofilament light chain (NfL) disclosed at 3 months or at 9 months. Neurologists and patients were blinded until the assigned visit. The primary outcome was the proportion with a very high-confidence (at least 90%) etiologic diagnosis at 3 months.
The effect was large. At 3 months, 56 of 112 participants (50.0%) in the earlier-disclosure arm had a very high-confidence diagnosis versus 5 of 108 (4.6%) in the delayed arm (P < .001). The gap appeared in each syndrome group: subjective decline 46.8% vs 7.5%, MCI 60.5% vs 3.3%, and mild dementia 40.7% vs 0%. Diagnoses changed more often with early disclosure (23.3% vs 7.0%, P = .002). Certainty was highest among people with low p-tau217 (61.3% vs 4.9%) but not the intermediate group (7.7% vs 11.1%), where the test is inherently less decisive.
Management differed too. Symptomatic Alzheimer drug treatment was started in 14.3% of the earlier arm versus 4.6% of the delayed arm (difference 9.7 points; 95% CI 1.8 to 17.8; P = .01). Cognitive stimulation referrals rose (45.5% vs 30.6%, P = .03). Planned repeat neuropsychological testing for diagnostic clarification fell (30.4% vs 52.8%; P < .001), and planned discharge to primary care roughly tripled (31.3% vs 12.0%; P < .001). Referral for disease-targeting therapy trials was numerically higher (4.5% vs 0%) but not significant (P = .06).
After the delayed group received its results at 9 months, certainty converged (58.0% vs 52.8%, P = .43), suggesting early disclosure speeds rather than permanently changes diagnostic confidence. The exception was MCI, where higher certainty persisted (74.3% vs 46.4%, P = .04). Importantly, earlier disclosure was not associated with greater anxiety, depressive symptoms, perceived stress, or worse quality of life, addressing a common worry about telling patients biomarker results.
Several caveats apply. This was a single-center study in one public Spanish hospital, with mostly White European participants and 9 months of follow-up. Outcomes were largely clinician-rated or intended management rather than long-term patient outcomes. Subgroup analyses were exploratory, and no multiplicity adjustment was made. The intermediate p-tau217 group and the high-NfL group were small, so conclusions there are limited.
Key Findings
- Very high-confidence (≥90%) etiologic diagnosis at 3 months: 50.0% (56/112) with earlier disclosure vs 4.6% (5/108) with delayed disclosure (P < .001)
- Symptomatic Alzheimer drug treatment was started in 14.3% vs 4.6% of patients (difference 9.7 points; 95% CI 1.8 to 17.8; P = .01)
- Planned follow-up neuropsychological reassessments fell from 52.8% to 30.4% (difference −22.4; P < .001)
- Planned discharge to primary care rose from 12.0% to 31.3% with earlier disclosure (difference 19.2; 95% CI 8.4 to 29.5; P < .001)
- Diagnostic reclassification between visits was more frequent with earlier disclosure: 23.3% vs 7.0% (P = .002)
- Certainty gain was seen with low (61.3% vs 4.9%) and high (45.9% vs 0%) p-tau217 likelihood, but not intermediate (7.7% vs 11.1%, P > .99)
- Earlier disclosure was not associated with increased anxiety, depression, perceived stress, or worse quality of life; certainty converged at 9 months (58.0% vs 52.8%) except in MCI (74.3% vs 46.4%, P = .04)
Methodology
PLASMAR was a prospective, single-center randomized clinical trial at Hospital del Mar (Barcelona) registered at ClinicalTrials.gov (NCT06246019) and reported per CONSORT. Of 384 screened, 220 patients with subjective cognitive decline, MCI, or mild dementia were randomized 1:1 by minimization (balancing sex, treating neurologist, and p-tau217 category) to plasma p-tau217/NfL disclosure at 3 or 9 months, with clinicians and patients blinded until then. The primary outcome was the proportion with ≥90% etiologic diagnostic certainty at 3 months, analyzed by intention-to-treat with Fisher exact tests; per-protocol analysis (n = 170) was consistent. Subgroup analyses were exploratory, tests were two-sided at P < .05, and no multiple-comparison adjustment was applied.
Study Limitations
The trial was conducted at a single public memory clinic in Spain with a predominantly White European population, limiting generalizability, and follow-up was only 9 months. Several outcomes were clinician-reported or intended management, and subgroup analyses were exploratory with no multiplicity adjustment and small sample sizes in some strata (intermediate p-tau217, high NfL). Conflict-of-interest disclosures were not included in the text available for this summary.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe:
