How You Handle Blood Samples Can Make or Break Alzheimer's Biomarker Tests

Blood-based biomarkers for Alzheimer's disease — including amyloid-beta (Aβ42, Aβ40), phosphorylated tau (pTau) isoforms, glial fibrillary acidic protein (GFAP), and neurofilament light (NfL) — have transformed how the disease is detected and monitored. As anti-amyloid therapies enter clinical use and blood tests move into primary care, the accuracy of these measurements depends not just on the assay itself but on what happens to the sample before analysis. The Global Biomarker Standardization Consortium set out to systematically quantify just how much routine handling decisions can distort results.

The study enrolled n=15 participants per experiment and tested seven key pre-analytical variables: primary collection tube type (EDTA, heparin lithium, serum, and specialty tubes like Sarstedt S-Monovette and BD P100), degree of hemolysis, centrifugation speed and temperature, delays before centrifugation, delays before storage/freezing, tube-to-tube transfers, and freeze-thaw cycles. Biomarkers were measured across four platforms — Simoa (single-molecule array), Lumipulse, MesoScale Discovery (MSD), and immunoprecipitation–mass spectrometry (IP-MS) — providing a cross-platform validation for the findings.

The most striking finding was the universal sensitivity to collection tube type: every biomarker varied by more than 10% across tube types, a threshold the consortium defined as clinically meaningful. Aβ42 and Aβ40 were the most vulnerable overall. Their levels declined by more than 10% under centrifugation delays as short as 60 minutes at room temperature (RT), and degradation was substantially steeper at RT compared to refrigeration at 2°C–8°C. Even brief storage delays before freezing caused significant Aβ losses. This is clinically critical because amyloid ratio (Aβ42/40) is now used in the AT(N) framework for AD diagnosis.

GFAP and NfL showed the opposite problem: their concentrations increased artificially by more than 10% when samples were stored at RT or −20°C rather than the recommended −80°C. This artifactual elevation could falsely suggest neurodegeneration or neuroinflammation where none exists. Freeze-thaw cycles also incrementally affected these markers. In contrast, pTau isoforms — particularly pTau217, which has emerged as a leading clinical candidate — demonstrated remarkable stability across the majority of pre-analytical conditions tested, making it more forgiving in real-world settings.

Based on these findings, the consortium produced an evidence-based, expert-consensus sample handling protocol. Key recommendations include: use EDTA tubes as the primary collection vessel; centrifuge within 60 minutes of collection at 2°C–8°C where possible; freeze aliquots at −80°C promptly; minimize freeze-thaw cycles; and avoid lithium heparin or serum tubes for Aβ measurement. The protocol also addresses centrifugation speed, acceptable hemolysis thresholds, and transfer tube practices. The authors acknowledge that while pTau217 tolerates variation well, harmonizing protocols across all biomarkers protects the integrity of multi-marker panels.

This work has direct implications for clinical laboratories now adopting AD blood tests following recent regulatory approvals, as well as for clinical trial sites globally. Inconsistent pre-analytical handling is a silent source of measurement error that can mimic disease progression or mask treatment response, undermining both diagnosis and drug development. The consortium's standardized protocol offers a practical, evidence-grounded roadmap to ensure that the revolution in blood-based AD diagnostics delivers on its promise.