Differences in Pharmacokinetics of Amphotericin B Lipid Formulations Despite Clinical Equivalence

SIR—Because of their lower toxicity, lipid formulations of amphotericin B (AmB) are frequently administered to critically ill patients who require treatment of suspected or proven systemic mycosis. Three lipid formulations that are used in clinical practice—liposomal AmB (AmBisome; NeXstar Pharmaceuticals), AmB colloidal dispersion (Amphotec; Sequus Pharmaceuticals) and AmB lipid complex (Abelcet; The Liposome Company)—appear to have impressive differences in their pharmacokinetics. For example, the volume of distribution for AmB lipid complex is reported to exceed that for liposomal AmB by a factor of 1000 [1]. The differences have been observed in numerous studies investigating time-concentration profiles of lipid formulations of AmB and have been ascribed to the different particle sizes and shapes of the 3 lipid formulations [2–6]. In contrast to these findings, the clinical efficacy of the 3 formulations has turned out to be approximately identical [7–17]. Thus, none of the 3 drugs is preferred in a recent Infectious Diseases Society of America guideline [18]. The striking contradiction between clinical equivalence and pharmacokinetic diversity has stimulated an engaged discussion in a previous issue of this journal [1, 19, 20]. As mentioned by Frothingham [1], total AmB levels have been determined in all pharmacokinetic studies of lipid formulations of AmB. We have developed a chromatographic method for the separate measurement of lipid-formulated AmB and AmB which has been liberated from its lipid binding [21]. The liberated fraction of AmB is bound to proteins in the plasma (protein-bound AmB) because AmB is insoluble in water. We applied our method to 3 groups of critically ill patients treated with conventional AmB (AmB deoxycholate, Amphotericin B; Bristol-Meyers-Squibb), liposomal AmB, and AmB colloidal dispersion, respectively [22]. The study was performed in accordance with the Declaration of Helsinki (1964) in the revised version of 1996 (Somerset West), the Guidelines of the International Conference of Harmonization (ICH), the Good Clinical Practice (GCP) Guidelines, and the Austrian drug law (AMG). Informed consent was obtained, when possible, or patients were informed about the scientific use of their clinical data after waking up from analgosedation. Pharmacokinetic characteristics were found to be different for liposomal AmB and AmB colloidal dispersion, in particular for their lipid-formulated fractions. Similar pharmacokinetics, however, were obtained for the liberated, protein-bound AmB fraction (see table 1), which resembled the pharmacokinetics of AmB deoxycholate [22]. The mechanisms of liberation of AmB from lipid binding are poorly understood. For liposomal AmB, it has been supposed that the bond between AmB and its lipid moiety is maintained in the circulation. Fungal or intrinsic phospholipases have been hypothesized to be involved in cleavage of AmB from its lipid moiety at the target site [23, 24]. Our data, however, suggest that, after infusion of liposomal AmB or AmB colloidal dispersion, significant amounts of AmB are liberated from the lipid moiety during circulation in the plasma. The pharmacokinetic properties of this liberated fraction are rather similar for both of the different formulations and resemble those of conventional AmB (see table 1). Pharmacokinetics of the lipid-formulated fractions of the liposomal AmB and AmB colloidal dispersion formulations, however, appear to be characteristic for the respective formulations. Pharmacokinetic data of lipoprotein-bound and lipid-formulated amphotericin B. Thus, the large differences in pharmacokinetic parameters between the lipid formulations can be attributed mainly to the lipid-formulated fractions. From a pharmacodynamic point of view, the equivalence in clinical efficacy of the different formulations supports a model that considers the liberated fraction of AmB to be the active form. The concentrations of lipid-bound AmB, however, probably do not directly influence the antimycotic activity of the drug. Thus, the contradiction between clinical equivalence and pharmacokinetic difference may be resolved by considering concentrations of active, liberated AmB in treatment with AmB lipid formulations.