Simultaneous quantification of linezolid and its metabolites (PNU-142300 and PNU-142586) in oral fluid and capillary blood by UPLC-MS/MS: method validation and clinical application using non-invasive sampling techniques

BACKGROUND: Therapeutic drug monitoring (TDM) can improve the safety and efficacy of the antibiotic linezolid (LZD), particularly in patients at risk of toxicity or with high pharmacokinetic variability. Although oral fluid (OF) and capillary blood collected through microsampling have emerged as less invasive alternatives to venous blood, the applicability of these matrices for quantifying LZD and its major metabolites-PNU-142300 and PNU-142586-has not been established. This study aimed to develop and validate a UPLC-MS/MS method for the simultaneous quantification of LZD, PNU-142300, and PNU-142586 in OF and capillary blood collected with volumetric absorptive microsampling (VAMS). METHODS: Method validation followed ICH M10 guidelines and included assessment of linearity, accuracy, precision, selectivity, matrix effects, hematocrit influence, dilution integrity, and short- and long-term stability. The method was further evaluated using paired clinical samples from 35 patients receiving LZD. RESULTS: The assay showed linearity from 0.1 to 5 µg/mL for all analytes in plasma, OF, and VAMS, with accuracy and precision within ICH M10 acceptance criteria, demonstrating analytical feasibility across all matrices. In the clinical phase, LZD showed strong correlations with plasma in both VAMS and OF, supporting their feasibility for linezolid monitoring. However, metabolites showed strong but systematically biased correlations in VAMS, indicating that correction factors will be required before routine clinical use. In OF, metabolites were rarely detectable, precluding their monitoring in this matrix. CONCLUSIONS: Overall, these findings demonstrate for the first time the feasibility of using VAMS and OF as non-invasive matrices for LZD TDM. Regarding feasibility, LZD concentrations in both matrices correlated strongly with venous plasma and the method met all ICH M10 validation criteria. Regarding current limitations, direct clinical substitution is not yet possible for metabolite monitoring: systematic underestimation in VAMS requires prospective validation of correction factors, and the near-absence of metabolites in OF limits this matrix to linezolid monitoring only. Larger studies are needed to validate correction strategies and establish matrix-specific therapeutic targets prior to clinical implementation.