Sample stability of forty-two analytes in plasma or serum pools after one to four repeated -80 °C freeze-thaw cycles

Knowledge of the pre-analytical stability is crucial when using frozen biobank samples for e.g. reference intervals studies. Such samples are sometimes subjected to multiple freeze-thaw cycles before analysis. There is a lack of comprehensive data on how repeated freezing and thawing affects the stability of many analytes in serum and plasma, particularly when compared to fresh samples. We studied the systematic effects of one to four -80 °C freeze-thaw cycles on alanine transaminase, albumin, alkaline phosphatase (ALP), alpha-1-antitrypsin, amylase, apolipoprotein A1 and B, aspartate transaminase, C-reactive protein (CRP), calcium, ceruloplasmin, chloride, creatine kinase (CK), creatinine, cystatin C, ferritin, gamma-glutamyl transferase, glucose, haptoglobin, high-density lipoprotein (HDL) cholesterol, immunoglobulin A, G and M, interleukin-6 (IL-6), iron, lipase, lipoprotein (a), low-density lipoprotein (LDL) cholesterol, magnesium, orosomucoid (alpha-1-acid glycoprotein), phosphate, potassium, sodium, total cholesterol, total protein, transferrin, transthyretin (prealbumin), triglyceride, urate and urea (carbamide) in serum pools and cardiac troponin T and N-terminal pro-brain natriuretic peptide (NT-proBNP) in serum and plasma pools. Mean relative changes after repeated freezing and thawing compared to samples analyzed fresh and samples frozen and thawed once were estimated. Most analytes were stable for up to four freeze-thaw cycles compared to fresh samples according to at least the minimum allowable bias (AB) based on biological variation. Analytes found to be unstable according to at least the desirable AB criterion after at least one of the freeze-thaw cycles were HDL cholesterol, chloride, and sodium.