Pre-Analytical Variables Influencing Stability of Blood-Based Biomarkers of Neuropathology

Background: Sample collection and preanalytical protocols may significantly impact the results of large-scale epidemiological studies incorporating blood-based biomarkers of neuropathology. Objective: To evaluate the stability and assay variability of several blood-based biomarkers of neuropathology for common preanalytical conditions. Methods: We collected serum and plasma samples from 41 participants and evaluated the effect of processing delay of up to 72 h when stored at 4 degrees C, three freeze-thaw cycles, and a combination of 48-h processing delay when stored at 4 degrees C and three freeze-thaw cycles on biomarker stability. Using the Simoa assay (Quanterix Inc.), we measured amyloid-beta 40 (A beta(40)), amyloid-beta 42 (A beta(42)), neurofilament light (NfL), glial fibrillary acidic protein (GFAP), and phosphorylated tau 181 (p-tau-181). Results: We found that A beta(40) and A beta(42) levels significantly decreased after a 24-h processing delay in both plasma and serum samples, and a single freeze-thaw cycle (p < 0.0001). Nevertheless, serum A beta(42/40) ratio remained stable with a processing delay up to 48 h while plasma A beta(42/40) ratio showed only small but significant increase with a delay up to 72 h. Both plasma and serum GFAP and NfL levels were only modestly affected by processing delay and freeze-thaw cycles. Plasma p-tau-181 levels notably increased with a 24-, 48-, and 72-h processing delay, but remained stable in serum. Intra-individual variation over two weeks was minimal for all biomarkers and their levels were substantially lower in serum when compared with plasma. Conclusion: These results suggest that standardizing preanalytical variables will allow robust measurements of biomarkers of neuropathology in population studies.