Plasma pharmacokinetics of [11 C]-flumazenil and its major metabolite [11 C]-flumazenil ‘acid’

[11C]-Flumazenil is a radiopharmaceutical that can be used to quantify benzodiazepine receptor concentrations and drug binding in the human brain using positron emission tomography (PET). In PET studies, arterial blood sampling is required to correct for labelled metabolites in plasma. The metabolite corrected arterial plasma curve of [11C]-flumazenil is used as the input function for the receptor pharmacokinetic model in clinical PET studies. The main metabolic pathway for flumazenil is conversion to flumazenil ‘acid’ by hepatic carboxylesterases. Interestingly, carboxylesterases are also involved in the bioconversion of other ester-type drugs such as cocaine, acetylsalicylic acid and many cholesterol synthesis inhibitors. Thus far, little is known about (genetic) differences in carboxylesterase activities. In principle, flumazenil may serve as a marker substrate for carboxylesterase phenotyping and predict metabolism for these kinds of drugs. The aim was to study interindividual differences in flumazenil clearance and flumazenil ‘acid’ formation in healthy volunteers and CNS patients. Healthy volunteers and CNS patients (n=25) participating in different PET protocols were included. During PET scanning, arterial blood samples were collected for ex vivo counting and metabolite analysis at 7 time-points. The plasma samples were analysed by reversed phase h.p.l.c. with radioactivity detection, as previously described [1]. The data were fitted by a commercially available and previously validated pharmacokinetic computer program (MW\Pharm, MediWare BV, Groningen, The Netherlands). The total body clearance (CL), volume of distribution (Vd) and the elimination rate constant (k10) were calculated for flumazenil pharmacokinetics. Parent flumazenil clearance from the plasma compartment was best fitted with a 2-compartment model (r2>0.98). The plasma clearance ranged from 48–139 l h−1. The flumazenil ‘acid' curve was fitted by an extravascular, 1-compartment model (r2>0.98). The rate constant of metabolite formation (km) ranged from 2.4–24 h−1, indicating pronounced interindividual differences in flumazenil ‘acid' formation. The total body clearance of this tracer dose of flumazenil is in the range of earlier reported pharmacological doses ([2]; Table 1). The estimated hepatic extraction ratio (EH) was 0.6–0.8 (assuming a hepatic blood flow of 60–90 l h−1). The rate of flumazenil ‘acid' formation correlated with the clearance of flumazenil, suggesting a major role of hepatic carboxylesterases in flumazenil clearance. Table 1. Pharmacokinetics of parent [11C]-flumazenil. Tracer dose Pharmacological dose Range Mean Clearance (l h−1) 48–139 76 ± 23 30–78 Vd (l kg−1) n.d. 0.78 (n.d.) 0.6–1.1 k10 (h−1) 1.7–25 4.6 ± 4.6 n.d. EH 0.6–0.8 n.d. 0.6 The plasma pharmacokinetics of [11C]-flumazenil show pronounced interindividual variation. The relative high hepatic extraction ratio implicates that the formation of flumazenil ‘acid' depends on the functional status of liver cells (carboxylesterase activity) and also on hepatic blood flow. Input curves of [11C]-flumazenil should be corrected for labelled metabolites on an individual basis. Interindividual differences in flumazenil handling may reflect the variation in drug metabolism of ester-type drugs.