Multi-site Analytical Evaluation of the Abbott ARCHITECT Tacrolimus Assay

Abstract

The objective of this study was to evaluate the analytical performance of the Abbott ARCHITECT Tacrolimus immunoassay. Proficiency panels and specimens from a population of organ transplant recipients were analyzed in 6 clinical laboratories in Europe and the United States, and the results were compared with other methods. The ARCHITECT assay requires a whole blood specimen pretreatment step with methanol/zinc sulfate to precipitate protein and extract the drug, followed by a 30-minute immunoassay using anti-tacrolimus antibody-coated paramagnetic microparticles and an acridinium-tacrolimus tracer. The assay was free from hematocrit interference in the range 25%-55% and from interference by extremes of cholesterol, triglycerides, bilirubin, total protein, and uric acid. The total percent of coefficient of variations of the assay were 4.9%-7.6% at 3 ng/mL, 2.9%-4.6% at 8.6 ng/mL, and 3.1%-8.2% at 15.5 ng/mL. Limit of detection was < or =0.5 ng/mL and limit of quantification (LOQ) ranged from 0.69 to 1.07 ng/mL across the 6 sites (based on the upper 95% confidence interval concentrations). The 2007 European Consensus Conference on Tacrolimus Optimization recommended the use of assay methods with an LOQ around 1 ng/mL, based upon the need to measure trough tacrolimus blood concentrations precisely down to 3 ng/mL during low-dose tacrolimus regimens. Tacrolimus International Proficiency Testing Scheme samples were measured by the ARCHITECT immunoassay at 5 sites and showed an average bias of -0.28 to +0.85 ng/mL versus IMx Tacrolimus II immunoassay historical values and -0.21 to +0.68 ng/mL versus liquid chromatography/tandem mass spectrometry (LC-MSMS) Tacrolimus historical values. Method comparison studies were performed with the ARCHITECT Tacrolimus immunoassay on patient specimens with the following results: ARCHITECT Tacrolimus assay versus the Abbott IMx Tacrolimus II immunoassay (4 sites) yielded average biases between -0.94 and +0.26 ng/mL; ARCHITECT assay versus the Dade Dimension Tacrolimus immunoassay (2 sites) yielded average biases of -0.46 and +0.11 ng/mL; and ARCHITECT assay versus LC-MSMS methods at 2 sites yielded average biases of +0.51 and +1.63 ng/mL. Spearman correlation coefficients were >/=0.90 on all method comparisons. The ARCHITECT Tacrolimus assay is a semiautomated, robust, and highly sensitive immunoassay, representing an alternative approach for laboratories not equipped with LC-MSMS, and meets the 1 ng/mL recommendation of LOQ by the European Consensus Conference on Tacrolimus Optimization.