Matrix fidelity in microsampling: Plasma-first LC-MS/MS quantification of mycophenolic acid and MPAG

Quantitative microsampling is increasingly used in bioanalysis; however, for plasma-referenced analytes, whole-blood microsampling may introduce matrix-dependent bias related to hematocrit (HCT) and blood-to-plasma analyte distribution. This study developed and validated a rapid liquid chromatography-tandem mass spectrometry (LC-MS/MS) method incorporating miniaturized micro-QuEChERS extraction for the simultaneous quantification of mycophenolic acid (MPA) and mycophenolic acid glucuronide (MPAG) in plasma, quantitative dried plasma spots (qDPS; Capitainer SEP10®), and quantitative dried blood spots (qDBS; Capitainer B®). Method validation, performed according to International Council for Harmonization (ICH) M10 and International Association of Therapeutic Drug Monitoring and Clinical Toxicology (IATDMCT) recommendations, demonstrated satisfactory selectivity, linearity, accuracy, precision, recovery, matrix-effect performance, stability, and incurred sample reanalysis across all matrices. In paired samples from pediatric kidney transplant recipients (n = 50), Passing-Bablok regression and Bland-Altman analysis showed close agreement between qDPS and conventional plasma, with low systematic deviation. For MPA, a small statistically significant proportional bias was observed, although most results remained within predefined clinical acceptance limits. In contrast, qDBS showed marked proportional bias and systematic underestimation compared with plasma. HCT- and regression-based correction strategies improved qDBS agreement but did not fully eliminate matrix-related discrepancies and require further external validation. These findings demonstrate that plasma-first microfluidic microsampling improves comparability with conventional plasma for MPA and MPAG by controlling matrix composition at the sampling stage. The proposed qDPS-based LC-MS/MS workflow provides a low-volume, matrix-aligned approach for plasma-referenced therapeutic drug monitoring and may support decentralized sampling in pediatric transplant recipients.