There are some indications from clinical studies (41,43) for aberrant cyclosporine metabolism resulting in formation of potentially toxic metabolites. When the activity of cytochrome P450 3A enzymes is low, more substrate is available for hypothetical alternative pathways of cyclosporine. There are several reasons for low P450 3A activity in a liver graft such as inter-individual genetic variability (43,49,84), cold ischemia and reperfusion damage, changes of the P450 activity during cholestasis (85) or other liver diseases (86), the influence of cytokines (87) and drug interactions such as inhibition or enzyme induction (88). Furthermore, low concentrations of cytochrome P450 3A influence the cyclosporine blood trough concentrations. The P450 3A concentration as estimated by the erythromycin breath test can be used to calculate the initial cyclosporine dose required to obtain cyclosporine blood trough concentrations in the therapeutic window (89). In vitro such alternative pathways comprising 3-methylcholanthrene-inducible (44,46,47) and/or ethinyl estradiol-inducible cytochrome P450 enzymes (48) could be identified and resulted in production of cyclized cyclosporine metabolites. The exact identification of the P450 enzymes involved requires metabolism of cyclosporine using reconstituted purified enzymes or single P450 enzymes expressed in cell lines. In addition, it remains to be clarified whether cyclosporine itself or its metabolite AM1 is the substrate for cyclization. Because cyclized metabolites have a low affinity to cyclophilin (58,59) they are mainly found in plasma. When more cyclized metabolites are formed primarily the concentration of cyclosporine metabolites in plasma increases. The free fraction of cyclosporine at 37 degrees C was found to be 1%-1.5% (90,91) of the cyclosporine concentration in blood. To date, nothing is known about the free fraction of cyclosporine metabolites. Because distribution characteristics of the cyclized metabolites in blood and urine are different from those of cyclosporine, it can be speculated that the free fraction of the cyclized metabolites is higher than that of cyclosporine. This might be reflected by a higher renal clearance resulting in relatively higher concentrations in urine compared with blood (61; Figure 3). If this is the case, a shift in the metabolite pattern with increased concentrations of cyclized metabolites will lead to an overproportional increase of the free fraction of cyclosporine metabolites. Although it is tempting to assume that cyclization is the alternative pathway explaining cyclosporine toxicity in patients with low concentrations of P450 3A enzymes in the liver (Figure 6), this has not yet been proven and will require not only quantification of P450 3A but of the complete P450 enzyme pattern in the liver in combination with characterization of the cyclosporine metabolite pattern by HPLC with special respect to the cyclized metabolites AM1c and AM1c9. Also, it is still unclear whether or not the cyclized metabolites contribute to cyclosporine toxicity. At least, it is unlikely that they are involved in covalent binding to macromolecules in the liver and kidney (44,71). In a clinical study using an HPLC method which allowed the specific quantification of 16 cyclosporine metabolites it was shown that the blood trough concentrations of the cyclized metabolite AM1c9 is elevated during early nephrotoxicity in liver graft recipients (82) and it was shown in an in vitro model that AM1c9 increases endothelin production and therefore might have a negative effect on renal hemodynamics.(ABSTRACT TRUNCATED)