Improvement of the pharmacokinetics and in vivo antibacterial efficacy of a novel type IIa topoisomerase inhibitor by formulation in liposomes

Abstract

Several useful properties of liposome-based formulations of various existing antibacterial drugs have been reported. These properties include lower MICs, improved pharmacokinetics, lower toxicity, selective distribution to infected tissues, and enhanced in vivo efficacy. Here we report in vivo studies of a liposomal formulation of a member of a novel class of antibacterial type II topoisomerase inhibitors, others of which have progressed to early phases of clinical trials. The free (i.e., nonliposomal) compound has broad-spectrum MICs but suboptimal pharmacokinetics in rats and mice, characterized by a high volume of distribution and rapid clearance. The liposomal formulation of the compound had essentially unchanged MICs but greatly reduced volume of distribution and clearance in rats and mice. In an in vivo mouse model of Staphylococcus aureus infection of one thigh, the liposomal compound localized preferentially to the infected thigh, whereas the free compound showed no preference for the infected versus the uninfected thigh. Most importantly, the liposomal compound had enhanced efficacy at clearing the infection compared with the free compound. Delivery of this class of compounds as liposomal formulations may offer clinical advantages compared with free compounds.

Fig 1

Structure of compound 1 (1-{2-[4-({[2,3-dihydro-1,4-dioxino(2,3-c)pyridin-7-yl]methyl}amino)-1-piperidinyl]ethyl}-1,2-dihydro-2-oxo-7-quinolinecarbonitrile). The two pK_a values of compound 1 are 5.2 and 8.2. The octanol-water (buffered at pH 7.4) partition coefficient (log D_7.4) is 0.68. Plasma protein was 51% and 72% bound with rat and mouse plasma, respectively (our unpublished observations).

Fig 2

Plasma pharmacokinetics of various liposomal formulations of compound 1 in rats and mice with intravenous administration. (A) Rat pharmacokinetics of 3 mg/kg free compound 1 in propylene glycol-ethanol-saline (5:4:1), 3 mg/kg of compound 1 in liposomes consisting of 1 mol DSPC:mol cholesterol, 2.8 mg/kg of compound 1 in liposomes consisting of 1 mol DPPC:mol cholesterol, and 2.5 mg/kg of compound 1 in liposomes consisting of 1 mol PEG-DSPE:10 mol HSPC:9 mol cholesterol. (B) Mouse pharmacokinetics of 10 mg/kg free compound 1 in a mixture of dextrose, povidone, and citric acid; 13.4 mg/kg of compound 1 in liposomes consisting of 1 mol DPPC:mol cholesterol; and 14.5 mg/kg of compound 1 in liposomes consisting of 1 mol DSPC:mol cholesterol. Data points and error bars represent averages and standard deviations of results from 2 or 3 animals. A summary of parameters calculated from these data are given in Table 1. The compound 1 concentration, hydrodynamic radius (R_h), and percent polydispersity index (%PI) for the liposome preparations in panel A were as follows: for 1 mol DSPC:mol cholesterol, 6.9 mM compound 1, R_h of 50 nm, and %PI of 0.24; for 1 mol DPPC:mol cholesterol, 6.8 mM compound 1, R_h of 52 nm, %PI of 0.25; and for PEG-DSPE:10 mol HSPC:9 mol cholesterol, 4.5 mM compound 1, R_h of 63 nm, and %PI of 0.36. The compound 1 concentration, R_h, and %PI for the liposome preparations in panel B were as follows: for 1 mol DPPC:mol cholesterol, 11.0 mM compound 1, R_h of 60 nm, and %PI of 0.30, and for 1 mol DSPC:mol cholesterol, 11.2 mM compound 1, R_h of 58 nm, and %PI of 0.30.

Fig 3

Concentrations of compound 1 in plasma (μg/ml) and in uninfected and infected thigh tissues (μg/g) of mice infected in one thigh with S. aureus. Mice were dosed intraperitoneally with 45 mg/kg/dose of free compound 1 in 5% (wt/vol) dextrose (A) or intravenously with 30 mg/kg/dose of compound 1 in DPPC:chol (1 mol:mol) liposomes (B) either once (solid lines) or twice at 12-h intervals. None of the free compound 1 measurements at 24 h for the uninfected thigh were above the 0.1-μg/ml limit of quantitation. The compound 1 concentrations after the second of two doses are shown with dashed lines. Symbols and error bars represent the means and standard deviations of measurements made with 3 mice. The compound 1 concentration, hydrodynamic radius (R_h), and percent polydispersity index (%PI) for the liposome preparation were as follows: 16 mM compound 1, R_h of 50 nm, and %PI of 0.49. Pharmacokinetic parameters for the plasma levels of compound 1 after the first dose of free or liposomal compound 1 are given in Table S1 in the supplemental material.

Fig 4

Effect of intraperitoneal free and intravenous liposomal compound 1 on growth of S. aureus in the thigh of neutropenic mice. (A) Mice were treated intraperitoneally with vehicle, consisting of 5% dextrose–2.5% Kollidon, or 1 to 4 doses spaced 3 h apart of 50 mg/kg free compound 1 in vehicle. Each group contained 8 mice. The difference in log CFU/g between the vehicle-treated mice and mice treated with a total dose of 200 mg/kg of compound 1 is statistically significant (P = 0.005 by 2-tailed t test). (B) Mice were treated with two doses at 12-h intervals of 5% dextrose (n = 5) or 45 mg/kg of free compound 1 dissolved in 5% (wt/vol) dextrose administered intraperitoneally (n = 3) or 30 mg/kg of compound 1 contained in DPPC:chol (1 mol:mol) liposomes administered intravenously (same preparation as in Fig. 3) (n = 3). There were 10 mice in the pretreatment control group. The difference in log CFU/g measured 24 h after the first dose between liposomal and free compound 1 is statistically significant (P = 0.001 by 2-tailed t test). The difference in log CFU/g between liposomal compound 1 measured 24 h after the first dose and the pretreatment control is statistically significant (P = 0.025 by 2-tailed t test). (C) Mice were treated with one intravenous dose of 28 mg/kg liposomal compound 1. S. aureus log CFU/g in the infected thigh was measured after 12 h (n = 3) and 24 h (n = 4). The pretreatment control group contained 10 mice. The difference in log CFU/g between the pretreatment control and 12 h posttreatment is statistically significant (P = 0.0004 by 2-tailed t test).