Challenged comparison of tacrolimus assays

Abstract

Monitoring tacrolimus trough concentrations is important for optimal immunosuppression in solid organ transplant recipients. Available assays generally correlate well with each other but little attention is given to patients in whom tacrolimus metabolite concentrations might be elevated, which could lead to artificially increased tacrolimus concentrations assessed by cross-reacting immunoassays. We addressed this hypothesis by investigating the correlation between four different assays (two immunoassays and two mass-spectrometry assays) in both a population with normal and a population with high dose requirements. Routine blood samples were collected in 37 control (CO) and 72 high dose patients (HD). Tacrolimus was measured with a CMIA, an ECLIA and two LCMS assays. Results were investigated using Deming regression analysis, Pearson correlation coefficients, Bland-Altman plots and by calculating bias. The CMIA demonstrated a positive bias of 23–26% compared with both LCMS assays. The correlation between CMIA and LCMS assays was good for the CO (r = 0.96) but less so for the HD group (r = 0.91). The ECLIA showed a positive bias of 11–13% compared with both LCMS assays. The correlation between ECLIA and LCMS assays was also good for the CO (r = 0.95) but again less for the HD group (r = 0.93). The correlation for both LCMS assays was excellent for either group (r > 0.99) with no bias. CMIA, ECLIA and LCMS assays for tacrolimus therefore correlate well for trough concentrations from solid organ transplant recipients. However, inter-assay differences exist, which seem more pronounced in patients who need a high dose of tacrolimus to reach a trough concentration in the therapeutic range.