Manual punch versus automated flow-through sample desorption for dried blood spot LC-MS/MS analysis of voriconazole

Abstract

Dried blood spot (DBS) sampling is a patient-friendly alternative for plasma sampling for the purpose of therapeutic drug monitoring (TDM). To speed up the analysis time, an automated flow-through desorption method of DBS samples may be beneficial. This article describes the cross-validation of a manual punch DBS method with an automated desorption (DBS autosampler, DBSA) method for the DBS analysis of the antifungal drug voriconazole, followed by cross-validation of both DBS methods with a plasma-based method, and an assessment of agreement between DBS/DBSA and regular plasma concentration measurements (gold standard) in samples from patients on voriconazole treatment. DBS and DBSA LC-MS/MS assays for voriconazole were validated according to the latest guidelines on bioanalytical method validation (FDA, EMA). Additional DBS-specific validation parameters included hematocrit effect and the influence of spot volume. Passing-Bablok regression and Bland-Altman plots were used to cross-validate the punch DBS, DBSA and plasma methods. The assessment of agreement between DBS/DBSA and plasma concentration measurements involved the performance of DBS/DBSA measurements to predict voriconazole plasma concentrations in patient samples. Both DBS methods complied with all validation parameters. Sample pre-processing time was reduced from 1.5 h to 3 min when using the DBSA. Cross-validation of both DBS methods showed a proportional bias and a correction factor was needed to interchange voriconazole concentrations of both DBS methods. Similarly, the punch DBS method required a factor to correct for proportional bias compared to the plasma method, but the DBSA and plasma assays showed no bias. Limits of agreement of the DBS/DBSA and plasma assays in Bland-Altman analysis were relatively wide, i.e. 0.75-1.28 for the DBS punch method versus plasma method and 0.57-1.38 for the DBSA versus plasma assay. Interpretation of DBS, DBSA and plasma samples in terms of concentrations in or outside of the voriconazole therapeutic range agreed in 82-86% of the cases. The variability in paired DBS/DBSA and plasma concentration measurements is considered high for TDM purposes and this limitation should be balanced against the advantages of DBS sampling of voriconazole and the speed of flow through desorption.