Pre-Analytical Variables Influencing Stability of Blood-Based Biomarkers of Neuropathology.
Author(s): Panikkar D, Vivek S, Crimmins E, Faul J, Langa KM, Thyagarajan B
Publication: J Alzheimers Dis, 2023, Vol. 95, Page 735-748
PubMed ID: 37574735 PubMed Review Paper? No
Purpose of Paper
This paper compared levels of amyloid-β 40 (Aβ40) and 42 (Aβ42), neurofilament light (NfL), glial fibrillary acidic protein (GFAP), and phosphorylated tau 181 (p-tau-181) between serum and plasma specimens, among plasma specimens obtained after a ≤72 h centrifugation delay at 4°C, among serum specimens that were stored on wet ice for 24 or 48 h prior to aliquoting, and among serum specimens that were freeze-thaw cycled between one and three times. The inter- and intra-individual variability of targeted markers in serum were also compared.
Conclusion of Paper
Levels of Aβ40, Aβ42, NfL, GFAP, and p-tau-181 were significantly higher in plasma than serum specimens, but the ratio of Aβ42 to Aβ40 was comparable in plasma and serum. When plasma and serum were compared, levels of both GFAP and p-tau 181 were moderately correlated, whereas the ratio of Aβ42 to Aβ40 and NfL levels were not. Levels of Aβ40 and Aβ42 decreased progressively in plasma with delayed centrifugation (24, 48, and 72 h at 4°C) and in serum when stored on ice prior to freezing (24 or 48 h) or when freeze-thaw cycled 1-3 times. The ratio of Aβ42 to Aβ40 increased progressively in plasma with delayed centrifugation at 4°C but was not significantly affected by the storage of serum prior to freezing or freeze-thaw cycling of serum. GFAP decreased progressively in plasma with delayed centrifugation. While GFAP levels were unaffected by the delayed freezing of serum, they were altered by freeze-thaw cycling serum specimens. Levels of NfL were unaffected in plasma by a 24 h centrifugation delay or in serum by storage prior to freezing, but declined in plasma when centrifugation was delayed by 48 h or 72 h and were affected in serum by freeze-thaw cycling. Levels of p-tau-181 were unaffected in plasma by a 24 h centrifugation delay, in serum by storage prior to freezing, and by freeze-thaw cycling of serum; p-tau-181 levels increased in plasma when centrifugation was delayed by 48 or 72 h. The inter- individual variability in all markers was greater than the intra-individual variability observed between the two time-points.
Studies
-
Study Purpose
This paper compared levels of Aβ40, Aβ42, NfL, GFAP, and p-tau-181 between serum and plasma specimens, among plasma specimens obtained after a ≤ 72 h centrifugation delay, among serum specimens that were stored on wet ice for 24 or 48 h prior to aliquoting, and among serum specimens that were freeze-thaw cycled one to three times. The inter- and intra-individual variability of targeted markers was also compared in serum. Blood was collected from 41 healthy volunteers after an overnight fast into four EDTA tubes (31 volunteers) and three double gel serum separator tubes. An additional SST was collected from twenty of the volunteers two weeks after the initial blood collection. Case-matched blood collected in EDTA tubes was stored at 4°C for 0, 24, 48, or 72 h before separation of plasma by centrifugation at 2000 g for 15 min. Plasma was then aliquoted and stored at -70°C. Blood in SST tubes was allowed to clot for 45 min at room temperature before serum separation by centrifugation at 1200 g for 15 min. Serum from 30 volunteers was aliquoted immediately or after 24 or 48 h storage on an ice pack and then frozen at -70°C. To investigate the effects of freeze-thawing, serum that was stored for 0 and 48 h before aliquoting was aliquoted and subjected to 1, 2, and 3 cycles of thawing at room temperature for 45 min and freezing for 24 h. Levels of Aβ40, Aβ42, NfL, and GFAP were quantified using the Neurology 4-Plex E Kit on an HD-Z analyzer, and levels of p-tau 181 were quantified using the Advantage V2.0 Kit on an HD-Z analyzer.
Summary of Findings:
Levels of Aβ40, Aβ42, NfL, GFAP, and p-tau-181 were significantly higher in plasma than serum (P≤0.0004, all), but the ratio of Aβ42 to Aβ40 was comparable in plasma and serum. While GFAP and p-tau 181 levels were moderately correlated between serum and plasma (r=0.41 and P=0.03, r=0.58 and P=0.001, respectively), the ratio of Aβ42 to Aβ40 and NfL levels were not correlated between serum and plasma. Plasma obtained after a 24, 48, and 72 h delay to centrifugation at 4°C had lower levels of Aβ40 (-15.72%, -35.46%, and -50.01%, respectively; P<0.0001, all), Aβ42 (-11.42%, -25.80%, and -38.26%, respectively; P<0.0001, all), and GFAP (-8.09%, -13.25%, and -15.25%, respectively; P≤0.0009, all); and a higher ratio of Aβ42 to Aβ40 (4.17%, 16.69%, and 30.60%, respectively; P≤0.006, all). While NfL and p-tau-181 levels in plasma were unaffected by a 24 h centrifugation delay at 4°C, plasma levels of NfL were lower when blood was stored for 48 or 72 h at 4°C before centrifugation (-8.26%, P=0.07 and -9.51%, P=0.006, respectively) whereas plasma levels of p-tau-181 were significantly higher (28.89%, P<0.001 and 59.90%, P<0.001, respectively). Serum stored for 24 and 48 h on ice before freezing had significantly lower levels of Aβ40 (P<0.0001, both) and Aβ42 (P<0.0001, both) and a non-significantly higher ratio of Aβ42 to Aβ40 (8.38% and 25.21%, respectively) compared to serum that was not stored; GFAP, NfL, and p-tau-181 levels were not significantly affected by storage of serum prior to freezing.
Immediately frozen serum that was freeze-thaw cycled once, twice, or three times had significantly lower levels of Aβ40 (-15.55%, -27.93%, and -32.95%, respectively; P<0.001, all) and Aβ42 (22.62%, 27.46%, and 30.51%, respectively; P<0.001, all) compared to immediately processed serum; the ratio of Aβ42 to Aβ40 and p-tau-181 levels were unaffected by freeze-thaw cycling of serum. GFAP and NfL levels were significantly higher after one freeze-thaw cycle (12.68%, P=0.02 and 14.57%, P=0.008, respectively), comparable to immediately processed serum after two cycles, and lower than immediately processed serum after three cycles (-7.58%, P=0.008 and -9.60%, P<0.0001, respectively). Similarly, the authors noted that storage of serum for 48 h prior to freeze-thaw cycling once, twice, or three times resulted in significantly lower levels of Aβ40 and Aβ42 (P<0.05, all) and a higher ratio of Aβ42 to Aβ40 (P<0.03, all), but it was unclear if the comparison was to the immediately processed specimen or to one with a 48 h processing delay. GFAP and NfL levels were significantly higher after one freeze-thaw cycle event (P=0.04 and P=0.002, respectively), comparable to baseline after two cycles, and lower than baseline after three cycles (P=0.003 and P=0.0007, respectively) but it was unclear if the comparison was to the immediately processed specimen or to one with a 48 h processing delay. The inter-individual variability in all evaluated markers was greater than the intra-individual variability observed between the two timepoints.
Biospecimens
Preservative Types
- Frozen
Diagnoses:
- Not specified
Platform:
Analyte Technology Platform Protein Clinical chemistry/auto analyzer Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Biospecimen Acquisition Time of biospecimen collection Initial collection
2 weeks later
Storage Freeze/thaw cycling 0 cycles
1 cycle
2 cycles
3 cycles
Storage Storage duration 0 h
24 h
48 h
72 h
Biospecimen Aliquots and Components Centrifugation Centrifugation delays investigated