Neurology · Medical Literature
Risk of False-Positives With Alzheimer Disease Blood Test
Abstract
This Medical News article discusses a recent study in JAMA Neurology about the real-world use of the first approved plasma test for biomarkers of Alzheimer disease pathology.
Clinical relevance
The FDA cleared the first plasma-based Alzheimer disease biomarker test — the Lumipulse G pTau217/β-Amyloid 1-42 Plasma Ratio — in May 2025 for symptomatic adults aged ≥55 years, representing a paradigm shift from invasive PET imaging and CSF lumbar puncture to a simple blood draw. This article in JAMA addresses the critical real-world gap between controlled trial performance and actual clinical deployment, specifically the risk of false-positive results that could lead to inappropriate diagnosis, unnecessary anti-amyloid therapy exposure, and significant patient psychological harm. As neurologists and primary care clinicians increasingly adopt this technology — particularly in the context of FDA-approved anti-amyloid immunotherapies (lecanemab, donanemab) that require biomarker confirmation — understanding the test's real-world limitations is of immediate clinical urgency.
Key findings
1. In the FDA pivotal multicenter clinical study (n=499 cognitively impaired adults), the Lumipulse G pTau217/β-Amyloid 1-42 Plasma Ratio demonstrated a positive concordance rate with amyloid PET or CSF of 91.7% and a negative concordance rate of 97.3%, implying an approximate false-positive rate of ~8.3% among test-positive results under controlled conditions. 2. In real-world implementation at the Mayo Clinic, investigators observed a shift toward a significantly higher rate of false-positive and indeterminate results compared with the controlled trial performance, with the FDA-cleared cutoffs producing notably lower diagnostic specificity than the research cutoffs previously validated by Palmqvist et al. 3. A large multicenter European validation study (n=1,767 symptomatic patients across 5 cohorts) using the Lumipulse immunoassay reported AUCs of 0.93–0.96, with positive predictive values of 89–95% in secondary care and accuracy of 85% in primary care — performance levels that may not fully translate to unselected real-world populations. 4. Less than 20% of patients in the FDA pivotal trial received an indeterminate result; however, real-world data suggest the intermediate zone may be larger when FDA-approved commercial cutoffs — which are lower than research-validated cutoffs — are applied, increasing clinical ambiguity. 5. Within-subject biological variation for plasma pTau217 has been quantified at 10.3% (95% CI: 9.2–11.7%), with between-subject variation of 21.1% (95% CI: 15.1–29.3%), indicating that a single measurement carries meaningful imprecision; three serial samples would be required to estimate an individual's true homeostatic value within 5% of the actual level. 6. A Class II FDA recall of specified Lumipulse reagent lots was initiated in December 2025 after reports of falsely elevated pTau217/Aβ42 ratios generating excessive positive or indeterminate classifications, directly confirming that manufacturing variability constitutes an additional real-world source of false positives beyond threshold-related issues. NOTE: Specific quantitative findings (e.g., exact false-positive rates, p-values, NNT/NNH) from the primary JAMA Medical News article itself were not retrievable from the abstract, which contained only the article description; the above data are drawn from the referenced JAMA Neurology study and closely related published evidence as identified through literature search.
Practice implications
Clinicians ordering the Lumipulse G pTau217/β-Amyloid 1-42 Plasma Ratio should restrict its use to symptomatic patients aged ≥55 years evaluated in a specialty care context, recognizing that the test's real-world false-positive rate may meaningfully exceed the 8.3% observed in the FDA pivotal trial — particularly when FDA-cleared commercial cutoffs are applied to populations differing from the validation cohort in disease prevalence, comorbidity burden (e.g., renal impairment, recent stroke), or assay lot. A positive result, especially one near the threshold or in the intermediate zone, should not be used as a standalone indication for anti-amyloid immunotherapy initiation; confirmatory testing with amyloid PET or CSF biomarkers should be strongly considered before irreversible treatment decisions, in line with current Alzheimer's Association clinical practice guidelines. Ordering clinicians should verify that the reagent lot used has not been subject to the ongoing FDA Class II recall and should counsel patients that a positive plasma biomarker result indicates increased probability of amyloid pathology — not a definitive diagnosis of Alzheimer disease — and that the clinical, functional, and neuroimaging context must be integrated before any diagnostic or therapeutic conclusion is reached.