The effect of freezing, thawing, and short- and long-term storage on serum thyrotropin, thyroid hormones, and thyroid autoantibodies: implications for analyzing samples stored in serum banks.

To the Editor: Data concerning the effects of freezing, thawing, and storage on serum thyroid-stimulating hormone (TSH) and thyroid hormones [free thyroxine (fT4) and free triiodothyronine (fT3)] are scarce (1–3), and we can find no data on their effects on thyroid autoantibodies [thyroid-peroxidase antibody (TPO-Ab) and thyroglobulin antibody (TG-Ab)]. We evaluated the effect of freezing, thawing, and storage of samples for up to 23 years on TSH, fT4, fT3, TPO-Ab, and TG-Ab concentrations in human serum (Architect i2000, Abbott Diagnostics). Whole blood samples for the Finnish Maternity Cohort (FMC) were collected without preservative from pregnant women, most of whom were in the 1st trimester of pregnancy. Each sample was stored in a single serum aliquot in a polypropylene cryovial (Nunc GmbH & Co.) at 25 °C. The ethics boards of the Finnish National Public Health Institute and Oulu University Hospital approved this study. The effect of freezing and thawing was evaluated by comparing frozen serum samples (n 50) with fresh samples (n 50) from FMC. Samples stored for 6 months vs 2–23 years at 25 °C (50 different samples collected from FMC at every time point, total n 645) were analyzed to evaluate the effect of storage time. Shortterm storage of up to 6 days at 4 °C was studied in fresh sera (n 8) from nonpregnant women. Details of the methods are available on request. After 0, 1, 3, and 6 days at 4 °C, no statistically significant effect was seen on any tests studied. There were no differences in TSH, fT4, TPO-Ab, or TG-Ab concentrations when 50 frozen and thawed serum samples were compared with 50 fresh serum samples. fT3 concentrations were significantly higher (Student t-test, P 0.001) in frozen samples but remained within reference intervals. We analyzed 645 samples from pregnant women with a mean gestational age of 11.9 weeks (range 6.7– 18.0) to evaluate the effect of storage time. Storage time had no effect on TSH or fT3 concentrations (Fig. 1). Concentrations of fT4 remained within reference intervals at all time points, although they differed significantly from baseline after 10, 12, and 20–23 years of storage (Fig. 1). Because fT4 concentrations were comparable between 0 and 14–18 years, the difference at 10 and 12 years might be the result of random variation. Storage time explained only 5.5% of the total variation in Table 1. Serum FLC measurements according to the dilutions.