Drug Metabolism by Cultured Human Hepatocytes: How Far are we from the In Vivo Reality?

The investigation of metabolism is an important milestone in the course of drug development. Drug metabolism is a determinant of drug pharmacokinetics variability in human beings. Fundamental to this are phenotypic differences, as well as genotypic differences, in the expression of the enzymes involved in drug metabolism. Genotypic variability is easy to identify by means of polymerase chain reaction-based or DNA chip-based methods, whereas phenotypic variability requires direct measurement of enzyme activities in liver, or, indirectly, measurement of the rate of metabolism of a given compound in vivo. There is a great deal of phenotypic variability in human beings, only a minor part being attributable to gene polymorphisms. Thus, enzyme activity measurements in a series of human livers, as well as in vivo studies with human volunteers, show that phenotypic variability is, by far, much greater than genotypic variability. In vitro models are currently used to investigate the hepatic metabolism of new compounds. Cultured human hepatocytes are considered to be the closest model to the human liver. However, the fact that hepatocytes are placed in a microenvironment that differs from that of the cells in the liver raises the question of to what extent drug metabolism variability observed in vitro actually reflects that in the liver in vivo. This issue has been examined by investigating the metabolism of the model compound, aceclofenac (an approved analgesic/anti-inflammatory drug), both in vitro and in vivo. Hepatocytes isolated from programmed liver biopsies were incubated with aceclofenac, and the metabolites formed were investigated by HPLC. The patients were given the drug during the course of clinical recovery, and the metabolites, largely present in urine, were analysed. In vitro and in vivo data from the same individual were compared. There was a good correlation between the in vitro and in vivo relative abundance of oxidised metabolites (4′-OH-aceclofenac + 4′-OH-diclofenac; Spearman's ρ = 0.855), and the hydrolysis of aceclofenac (diclofenac + 4′-OH-aceclofenac + 4′-OH-diclofenac; ρ = 0.691), while the conjugation of the drug in vitro was somewhat lower than in vivo. Globally, the metabolism of aceclofenac in vitro correlated with the amount of metabolites excreted in urine after 16 hours (ρ = 0.95). Overall, although differing among assays, the in vitro/in vivo metabolism data for each patient were surprisingly similar. Thus, the variability observed in vitro appears to reflect genuine phenotypic variability among the donors.