![]() |
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Received for publication May 3, 2005.
Revised September 15, 2005.
Accepted for publication September 16, 2005.
Abstract Predictions of intrinsic clearance (CLint) from human liver microsomes often underestimate in vivo observations. In this study, a series of five benzodiazepines were used as prototypic CYP3A4 substrates in order to investigate the prediction of clearance from the less studied alternative in vitro system, cryopreserved human hepatocytes. Formation of the two major metabolites from midazolam, triazolam, diazepam, flunitrazepam and alprazolam was measured in cryopreserved human hepatocytes from five donors; the kinetics were characterised and CLint values determined and scaled to predict CLint in vivo. At least one of the two major pathways of metabolic clearance of each benzodiazepine was characterised by autoactivation in hepatocytes; the extent to which this occurred varied depending on substrate and liver (up to 8-fold). Heteroactivation by testosterone of these pathways was also observed (up to 6-fold). The maximum autoactivated clearance was used to predict in vivo CLint (1.6 - 138ml/min/kg) which closely agreed with values previously obtained using human liver microsomes. Comparison with in vivo CLint indicates that cryopreserved human hepatocytes systematically under-predict CLint. When three previous studies (documenting CLint for substrates of various enzymes in cryopreserved human hepatocytes using drug depletion-time profiles) were considered as well, the combined data show a consistent underprediction of 5.6-fold. Collectively it is demonstrated that the predicted hepatic intrinsic clearances from cryopreserved hepatocytes show an excellent rank order with in vivo findings but are systematically under-predicting the in vivo value.
Key words:
drug clearance, enzyme kinetics, hepatic elimination, hepatocytes, human CYP enzymes, in vitro-in vivo prediction, in vitro-in vivo scaling