Applications of human microdosing with accelerator mass spectrometry: assessment of ability to predict drug-drug interactions and determine the pharmacokinetics of enantiomers

In this thesis, new applications of microdosing were explored in two clinical trials. Methods were developed for the separation and quantification of 14C-labelled analytes in human plasma using two-dimensional HPLC and accelerator mass spectrometry (AMS). Caffeine, midazolam, tolbutamide and fexofenadine were quantified in plasma after administration of a 14C-labelled cassette microdose to human volunteers via a HPLC-AMS recovery constant method. Mean accuracy for all analytes was within 13% of the measured plasma concentration with precision of <20% CV, meeting recommended acceptance criteria for HPLC-AMS assays. Complete resolution of each analyte was demonstrated by two-dimensional HPLC. Pharmacokinetic data obtained after cassette microdose administration were in close agreement with those previously obtained after administration of therapeutic doses. Co-administration of the cassette microdose with known inhibitors of metabolism enzymes and transporters resulted in a significant (p<0.01) increase in the area under the concentration-time curve from time zero to infinity (AUC0-∞) for caffeine (x8.2), midazolam (x11.7), fexofenadine (x3.2) and tolbutamide (x1.8, p<0.05). Administration of a combined 11C and 14C-labelled verapamil microdose allowed distribution in the brain to be monitored by PET imaging, while simultaneously obtaining plasma pharmacokinetics by AMS. The separation of 14C-verapamil by two-dimensional HPLC and AMS analysis resulted in the individual pharmacokinetics of R- and S-verapamil being consistent with those reported after therapeutic doses. In addition, a significant difference in pharmacokinetic data obtained for the two enantiomers clearly showed the preferential clearance of S-verapamil. Data were accurate within 12% of the true value with precision of <18% CV. Pharmacokinetic data obtained after PET analysis were consistent with those obtained during AMS analysis, proving the concept of combining the two techniques in clinical studies and enabling maximum information to be achieved from one single study.