Comparison of a Point-of-Care Testing with Enzyme-Multiplied Immunoassay Technique and Liquid Chromatography Combined With Tandem Mass Spectrometry Methods for Therapeutic Drug Monitoring of Mycophenolic Acid: A Preliminary Study
Hong Zhou,Hongping Xiang,Jie Cai,Yirong Wang,Min Zhang,Yong Han,Yu Zhang
DOI: https://doi.org/10.1097/FTD.0000000000000861
2021-10-01
Abstract:Background: For mycophenolic acid (MPA), therapeutic drug monitoring is an essential tool for dosage optimization in transplant recipients and autoimmune diseases. In China, a new commercial kit using an immunochromatographic assay (FICA) with a point-of-care testing system was approved for therapeutic drug monitoring of MPA. However, corroboration between FICA and clinically used assays remains unknown. The authors evaluated MPA concentrations in heart transplant recipients obtained by FICA, high-performance liquid chromatography combined with tandem mass spectrometry (LC-MS/MS), and enzyme-multiplied immunoassay technique (EMIT). Methods: Nine heart transplant recipients administered a single mycophenolate mofetil (MMF) dose, and 4 administered multiple MMF doses were enrolled. MPA samples were collected before administration, and after 0.5, 1, 1.5, 2, 4, 6, 8, 10, and 12 hours, and assessed by 2 immunoassays (EMIT and FICA) and LC-MS/MS. Consistency between methods was evaluated using Passing-Bablok regression and Bland-Altman analysis. Results: For Passing-Bablok regression between FICA and LC-MS/MS, FICA = 0.784 LC-MS/MS + 0.360 (95% CI slope: 0.739 to 0.829, 95% CI intercept: 0.174-0.545). Regardless of a significant observed correlation coefficient (R2 = 0.9126), statistical analyses revealed a significant difference between FICA and the reference LC-MS/MS method. The mean absolute bias was 0.69 mcg/mL between FICA and LC-MS/MS. Bland-Altman plots showed a mean bias of -0.23 mcg/mL (±1.96 SD, -2.19 to 1.72 mcg/mL) and average relative bias of 14.73% (±1.96 SD, -67.91% to 97.37%) between FICA and LC-MS/MS. Unsatisfactory consistency was observed between EMIT and LC-MS/MS, and FICA and EMIT. Differences between pharmacokinetic parameters after a single or 7 days of MMF administration, by LC-MS/MS and FICA, were not statistically significant. Conclusions: The consistency of the new FICA using a point-of-care testing device with LC-MS/MS and EMIT was inadequate, and the accuracy of EMIT and LC-MS/MS was inappropriate. Clinicians should be informed when switching MPA detection methods to avoid misleading results.