Stability of N-Terminal Pro-Brain Natriuretic Peptide after Storage Frozen for One Year and after Multiple Freeze-Thaw Cycles

Cardiac natriuretic peptides are of interest for their potential role in assisting in the diagnosis, prognosis, and monitoring of left ventricular dysfunction and congestive heart failure (1). An electrochemiluminescence immunoassay for the measurement of N-terminal pro-brain natriuretic peptide (NT-proBNP) on the Elecsys® immunoassay analyzer platform (Roche Diagnostics Corporation) has recently received Food and Drug Administration clearance to aid in the diagnosis of congestive heart failure (2). We report here on the stability of the NT-proBNP analyte when measured with the Roche Diagnostic NT-proBNP assay after specimen storage at −80 °C for longer than 1 year and after five freeze-thaw (−80 and 22 °C) cycles. Twenty-five samples were collected from healthy and heart failure clinic adult volunteers under informed consent according to the policies of the Washington University School of Medicine’s Institutional Review Board. All glass collection tubes were from Becton Dickinson. Lithium-heparin plasma (prod. no. 367686) was collected from all participants (n = 25). Serum was collected in either 10-mL Vacutainer® SST (prod. no. 366510) or 10-mL silicone-coated serum tubes (prod. no. 366430). EDTA plasma was collected in either 3-mL K2EDTA tubes (prod no. 367856,) or 7-mL K3EDTA tubes (prod. no 366450). After processing of samples, baseline values were measured with the Roche Diagnostics NT-proBNP assay on an Elecsys 2010 immunoassay analyzer at the Core Laboratory for Clinical Studies at Washington University School of Medicine. NT-proBNP concentrations in the samples ranged from 26 to 6838 ng/L. We then dispensed 1-mL aliquots into 2-mL polypropylene screw-cap vials (Sarstedt Inc.). After storage at −80 °C for longer than 1 year (mean time, 419 days), samples were removed and allowed to thaw at room temperature, and NT-proBNP concentrations were determined by single measurements. For the freeze-thaw study, serum was collected in silicone-coated serum tubes from four individuals. Baseline values were calculated as the mean of triplicate measurements. The NT-proBNP concentrations in the samples ranged from 85 to 5828 ng/L. We placed 1 mL of serum in a 2-mL polypropylene screw-cap vial and immediately stored it at −80 °C Samples were frozen for at least 24 h between measurements, at which time the sample was removed from −80 °C and thawed to room temperature, and the NT-proBNP concentration was measured in triplicate. The effect of long-term storage was analyzed statistically by the nonparametric Wilcoxon signed-ranks test using Microsoft® Excel 2000 with the Analyze-It® plug-in statistics module. When we tested the hypothesis that the original value would not be equal to the repeat value, the difference was not statistically significant under any of the blood collection conditions. The median (range) NT-proBNP concentrations (ng/L) for the tested matrices were as follows: Lithium-heparin plasma: original measurement, 607 (29–6838) ng/L; repeat, 585 (26–6373) ng/L SST serum: original measurement, 469 (31–2702) ng/L; repeat, 478 (38–2715) ng/L Siliconized serum: original measurement, 1043 (29–6791) ng/L; repeat, 749 (26–6592) ng/L K2EDTA plasma: original measurement, 1000 (26–6820) ng/L; repeat, 707 (26–6550) ng/L K3EDTA plasma: original measurement, 461 (27–2511) ng/L; repeat, 459 (38–2537) ng/L As seen in Fig. 1 , there was no observable increasing or decreasing trend in the sample concentrations over the course of the study. Plot of the percentage difference between the repeat value and the original value. The percentage difference was calculated as: (repeat value − original value)/original value × 100. The effect of the freeze-thaw process was evaluated by calculating the percentage of change (repeat value − original value)/original value × 100 from baseline. The range of change observed over all five cycles and across all samples was −5.4% to 7.2%. The mean change from fresh to frozen for the four samples at cycle five was −1.6%. We have demonstrated that NT-proBNP is stable for at least 1 year when stored at −80 °C in several different serum and plasma conditions. Because the Wilcoxon P values were >0.05, we cannot claim statistical significance, and we conclude that the data provide no evidence for a significant difference in scores between the original and repeat values. In addition, the peptide fragment remains nearly 100% immunoreactive in the Roche Diagnostics NT-proBNP assay despite at least five freeze-thaw cycles. In light of recent publications on the clinical utility of NT-proBNP as a biomarker for the evaluation of heart failure (3)(4), researchers and clinicians may have interest in the retrospective analysis of serum or plasma samples for NT-proBNP analysis. The above information should be useful in evaluating samples for NT-proBNP when stored under the conditions described here.