Abstract Background Measurement of plasma oxalate is important in patients with renal failure as oxalate may build up due to impaired excretion and lead to organ damage due to precipitation of calcium oxalate crystals. However, the measurement of plasma oxalate is challenging due to the low concentrations present in individuals with healthy kidney function and the pre-analytical steps needed to prevent erroneous oxalate results due to conversion from ascorbate. In order to provide improved turn-around-time for plasma oxalate testing, our objective was to develop and validate a method to measure plasma oxalate on the Roche cobas c501. Methods Samples used in the validation study included both fresh collections and residual plasma from specimens collected in lithium heparin, no-gel tubes. Samples from fresh collections were kept on ice prior to centrifugation, and then frozen if not analyzed immediately. Thawed plasma was acidified by addition of 12.1 M HCl to a pH between 2 and 3 and mixed thoroughly by vortexing. One milliliter of the acidified plasma was added to a Centrifree filter (Millipore). The samples were then centrifuged for 90 min at 1500 × g 20°C in a fixed angle rotor to remove plasma proteins. The resulting filtrate was assayed on the Roche cobas c501. The assay (Trinity Biotech) uses oxalate oxidase to produce an indamine dye through the production of hydrogen peroxide from oxalate in the sample, which is measured spectrophotometrically at 600 nm at a single time point. The analytical measuring range (AMR) was evaluated by performing serial dilutions of a high sample. Interferences from hemolysis, icterus, and lipemia were assessed at both low and high oxalate concentrations. Precision was evaluated using two QC concentrations, and patient samples with low and high concentrations of oxalate. Samples were tested ten times in a single day, or twice per day for ten days to estimate intra- and inter-precision, respectively. Spike recovery studies were used to assess method accuracy. Lastly, reference intervals were verified by assaying samples from twenty healthy volunteers. Total allowable error was ±30% or 2 µmol/L with a maximum 15%CV. Results The assay was validated to be linear across an AMR of 2 to 30 µmol/L, with an average recovery of 105% across five concentrations. The total assay imprecision was <15%. Initial spike recovery studies across the AMR showed an average recovery of 99.4% (n = 12), with a range of 76.6 to 132.4%. In a mixing study, hemolyzed and lipemic sample recovery was on average 82.1% and 76.1% respectively, which was low, but acceptable. Samples mixed with icteric specimens demonstrated unacceptably low recovery (average of 64.1%). The reference interval of <2 µmol/L was successfully verified as all twenty healthy volunteers had results <2 µmol/L. Conclusion Initial validation studies support the clinical utility of the newly developed method. Plasma oxalate methods are documented to have significant inter-method variability and variable recovery, which is hypothesized to be due to variation at the filtration step. This semi-automated assay will require less technologist time and will reduce the possibility for human error affecting the result.
Read full abstract