Long-term storage stability of plasma TNF-α and IL-6 concentrations of psychiatric emergency patients: The Signature Biobank

Cytokines are increasingly incorporated in psychiatric research. While biobanking provides advantages for different study designs, long-term storage could degrade cytokine quality and introduce bias. To date, reports on cytokine stability are based on small sample sizes (n < 20) and do not address long-term (4+ years) storage effects. This article reports two studies evaluating IL-6 and TNF-α measurements in human plasma samples biobanked for long periods. In the first study, with samples stored between 5 and 139 months (11.6 years), IL-6 concentrations were not significantly correlated with storage length in all available baseline data of the Signature Biobank (n[IL-6] = 1206, n[TNF-α] = 1223). TNF-α correlated negatively with storage length (r = -0.217; p < 0.001) indicating a decrease of concentrations with longer storage. In the second study, IL-6 and TNF-α concentrations were measured twice in the same plasma samples of 50 psychiatric participants from the Signature Biobank stored between analyses from 32 to 45 months. We assessed if storage effects differed between analytes. In IL-6, we observe a relatively good stability in samples stored up to 6 years. For TNF-α, we observed only a moderate stability of measurements (rTNF-α = 0.59; rIL-6 = 0.71) with linear decrease over time. As a potential solution, a corrective equation extracted from Study 1 was applied to TNF-α in the Study 2 sample; however, this did not improve correlation coefficients but might be useful in other settings. Integrating samples' age in statistical analyses and/or more systematic quality controls could mitigate degradation process.