Abstract
Recent published papers have highlighted the importance of preanalytical and analytical factors with regard to interpretation of cardiac troponin (cTn). 1,2 Specifically, a decrease in cTn concentration may occur due to stability issues 1 or the presence of an interference (i.e. haemolysis) 2 which may be inappropriately interpreted as a clinically significant change. 3,4 Thus, it is imperative that preanalytical and analytical factors are documented for each cTn assay such that the laboratory can properly restrict retrospective ‘add-on’ testing or measurement in samples with interferences present. Recently, Chenevier-Gobeaux et al. 2 have reported on factors affecting analytical performance of high-sensitivity cTnT (hs-cTnT); however, stability was not assessed. Surprisingly, limited data exist in this area with respect to freeze–thaw effects on hs-cTnT concentrations or on stability in different matrices, with a report earlier this year indicating that hs-cTnT concentrations are stable up to 24 h at room temperature in serum. 5 Therefore, we performed a study assessing stability and the freeze–thaw effect on hs-cTnT concentrations in different matrices and over different timeframes.
To assess stability, we constructed an EDTA plasma pool with an average hs-cTnT concentration of 50 (standard deviation [SD] = 1; triplicate measurement on the Roche E modular platform; Roche Diagnostics, Laval, QC, Canada) ng/L; representing a clinically relevant concentration. 6 This pool was split into two aliquots, with the first aliquot refrigerated (2–8°C) and the second frozen at −20°C. After 24 h, the refrigerated aliquot hs-cTnT measured concentration was 50 (1) ng/L and the frozen aliquot after one thaw measured concentration was also 50 (1) ng/L. The thawed aliquot was subjected to two additional freeze–thaws (one day and one week later) with no decrease in hs-cTnT concentrations observed (average hs-cTnT concentrations: 51 [1] ng/L and 52 [1] ng/L, respectively).
Next, we assessed the stability of hs-cTnT concentrations when left on cells. For this experiment, we obtained patient samples (lithium–heparin plasma centrifuged and not separated from cells, duplicate measurement; sample 1, hs-cTnT concentration = 108 [1] ng/L; sample 2, hs-cTnT concentration = 777 [4] ng/L). These same samples were measured for hs-cTnT daily over one week with the samples refrigerated (2–8°C) between analyses. The concentrations from day 1 to day 7 ranged from 109 to 106 ng/L and 773 to 748 ng/L for samples 1 and 2, respectively, corresponding to a percent difference of −1.9 and −3.7% at day 7 as compared with baseline.
From these analyses, it would appear that hs-cTnT concentrations are stable when refrigerated up to one week and after three freeze–thaw cycles. Thus, ‘add-ons’ for hs-cTnT measurements on refrigerated samples may be suitable and is in contrast to the Siemens cTnI-Ultra assay (Siemens, Deerfield, IL, USA). 2 These data highlight the fact that stability is a characteristic of the analyte and assay and, as such, should be determined for all cTn assays. This is especially important when considering if a change in concentration of cTn is reflective of an acute injury (i.e. acute myocardial infarction, AMI); as opposed to either a preanalytical or even an analytical issue. These present data are particularly relevant for interpreting retrospective studies employing the hs-cTnT assay; especially studies assessing change criteria for the diagnosis of AMI. 4
DECLARATIONS