Résumé
Background
Pre‐analytical sample handling might affect the results of Alzheimer’s blood‐based biomarkers. In this collaborative study of the SABB‐GBSC working group guided by the Alzheimer’s Association, we elucidate the effects of the most common sample handling variations on blood‐based amyloid and non‐amyloid biomarkers.
Method
Freshly collected blood was differentially treated, to create sample sets (n=10 per condition). We investigated the effect of collection tube type, and of EDTA plasma handling variations: delayed centrifugation, centrifugation temperature, aliquot volume, delayed storage, and freeze‐thawing. Frozen aliquots were shipped to participating labs, to measure amyloid beta (Abeta) peptides (with different technologies: C2N LS‐mass spectrometry (MS), Shimadzu MALID‐MS, N4PE and N3PA Simoa, Euroimmun and Araclon ELISA), Abeta oligomerization tendency (OAβ; PeopleBio), glial fibrillary acidic protein (GFAP; Quanterix), neurofilament light (NfL; Quanterix), total tau (tTau; Quanterix) and phosphorylated tau181 (pTau181; Eli Lilly Simoa assay). Data analysis was descriptive.
Result
Using other sample types than EDTA plasma resulted in different absolute levels for most of the measured biomarkers (figures 1‐4), with generally higher levels in lithium‐heparin plasma, lower levels in sodium‐citrate plasma, and variable but comparable levels in serum. Independent of the technology used, when EDTA plasma was kept at room temperature pre‐ or post‐centrifugation, Abeta42 and Abeta40 levels remained stable for up to 3 hours, while at 24 hours levels were decreased. When kept at 2‐8°C, Abeta levels were stable for up to 24 hours delayed centrifugation or storage (figure 1 and 2). The Abeta42/40 ratio mitigated some of the variability noted in Abeta42 and Abeta40 levels for some of the technologies and conditions, but not for all (figure 3). Abeta42 and Abeta40 levels were robust to differences in tube filling, short‐term ‐20°C storage, and up to 4 freeze‐thaw cycles (figure 1 and 2). While the other biomarkers OAβ, GFAP, NfL and pTau181 were mostly resistant to pre‐analytical sample handling variations, tTau levels were affected by delayed centrifugation and by centrifugation temperature (figure 4).
Conclusion
Based on the current findings, we construct a standardized operating procedure for pre‐analytical blood sample handling, that will facilitate swift introduction of blood‐based biomarkers into research and clinical settings.