Evaluation of a batched-extraction method for measurement of sirolimus, tacrolimus, and cyclosporine on the Architect i2000SR

Abstract

BackgroundManual extraction of immunosuppressants is required before measurement on the Architect immunoassay analyzer. The individual extraction of samples places clinical laboratory staff at risk for ergonomic injury. We evaluated the analytical performance of a batched extraction method for measuring sirolimus, tacrolimus, and cyclosporine using the Architect i2000SR. MethodsResidual whole blood samples from patients receiving immunosuppressant therapy were used for evaluation. The analytical evaluation included imprecision, linearity, and method comparison. Technologist-to-technologist variation was also assessed. ResultsTotal imprecision ranged from 2.58 to 3.13% for sirolimus, 2.70–3.77% for tacrolimus, and 7.82–12.41% for cyclosporine. Linearity was verified from 0.44–19.49 μg/l for sirolimus, 0.05–26.15 μg/l for tacrolimus, and 0.15–991.55 μg/l for cyclosporine. Deming regression analysis showed slope and intercept were not significant for either technologist-to-technologist comparison or for batched vs. individual processing comparison. Bland-Altman analysis of individual vs. batched processing revealed a mean bias of 1.29% (LLOA: −14.63%, ULOA: 17.21%) for sirolimus, 2.07% (LLOA: −10.87%, ULOA: 15%) for tacrolimus, and −1.56% (LLOA: −20.05%, ULOA: 16.94%) for cyclosporine. The values were not significantly different from the bias and LLOAs observed for technologist-to-technologist comparison. ConclusionsThe imprecision and linearity of batched methods met analytical goals. The batched method also correlated well with the individual extraction methods. The ULOA and LLOA for all drugs tested exceeded a TAE or ± 15%. However, similar range of differences was observed between technologists, suggesting that batch processing did not increase or reduce variability due to manual preparation steps.