Evaluation of microparticle enzyme immunoassay against HPLC-mass spectrometry for the determination of whole-blood tacrolimus in heart- and lung-transplant recipients

Abstract

Objectives: Tacrolimus is an immunosuppressant drug with a narrow therapeutic window and thus requires therapeutic drug monitoring. This study evaluates the suitability of the second-generation microparticle enzyme immunoassay (MEIA II) against a specific method, high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS), for the measurement of tacrolimus in both heart- and lung-transplant groups. A secondary objective was to investigate the effect of tacrolimus concentration on MEIA II measurement. Methods: The HPLC-MS assay was conducted as per our reported method and MEIA II performed according to manufacturer’s instructions. Quality-control samples at 5, 11, and 22 μg/L were run in each batch to ensure assay integrity in both methods. Multiple trough samples from 18 heart patients ( n = 126) and 17 lung patients ( n = 203) were analyzed. Results: The inter-batch imprecision and analytical recovery over the quality-control range by HPLC-MS ( n = 12) was <6% and 98.2% to 104%, respectively, and by MEIA II ( n = 16) <15% and 92.0% to 99.1%, respectively. The mean overestimation by MEIA II between the two methods for heart- and lung-transplant patient samples was found to be 9.9% (range: −37.4–45.4%) and 13.2% (range: −29.2–64.3%), respectively. Stratification of these data based on the tacrolimus concentration determined by MEIA II, yielded no statistically significant differences in bias between concentration subgroups within the clinically relevant range ( p > 0.4). However, a statistically significant difference was detected between the highest concentration subgroup (>20.0 μg/L) and lower concentration subgroups in both transplant populations ( p < 0.05). Conclusions: This study suggests that where HPLC-MS is not available, MEIA II may be suitable for the therapeutic drug monitoring of tacrolimus in heart- and lung-transplant recipients. However, the clinical importance of the observed mean bias, considering the wide range in overestimation in heart- and lung-transplant patient samples, is yet to be determined.