Selection of Pegylated Nanoliposomes as the Basis of Doxil
Each of the laboratories working on long circulating liposomes described above has its publications and patents on its unique liposome formulations. In the early 1990s, it became evident that for various scientific and practical reasons (availability, cost, species specificity, etc.) described elsewhere (Woodle, 1993; Allen, 1995), the GMl-ganglioside-based formulation was excluded. At LTI, HPI and PEG-DSPE remained in the race. In order for LTI to decide which of the two lipids will be the one to use in evaluating the novel formulation in human, we performed in 1991 a critical comparative PK experiment in Beagle dogs (Gabizon et al., 1993). Dog is a preferred animal due to its much larger plasma volume (~500 mL), which resembles much better that of human beings in studying the effect of dilution-induced drug release (Gabizon et al., 1991; Amselem et al., 1993a) than small rodents with their very small plasma volume (and therefore almost no dilution).
The Beagle dog doxorubicin PK study clearly demonstrated that although both liposomal formulations were much superior to F-DOX.
The PEG-DSPE formulation was better, as it showed much slower plasma clearance than the HPI formulation.
Figure 15.5 A cartoon showing a comparison between conventional and sterically stabilized (pegylated) liposomes (SSL, Stealth liposomes). (Courtesy of the late D. Lasic.) The cartoon shows lack of insertion of opsonins into the membrane of Stealth liposomes.
Using mice peritoneal macrophages (obtained from the ascitic fluid of mice treated with thioglycolate), in vitro, Doxil showed 40% of the uptake of liposomes of similar size and lipid composition but lacking 5 mole% PEG-DSPE (Emanuel et al., 1996a). This reduction in macrophage uptake is in direct correlation with the increase in plasma circulation time. The increase in mole% of PEG-DSPE can further reduce macrophage uptake and probably increase circulation time (Emanuel et ah, 1996a, 1996b). For more details on the effect of PEG-DSPE on nano-liposomes, see Garbuzenko etal. (2005).
15.4.6 Remote Loading of Doxorubicin into Nano
Remote Loading of Doxorubicin into Nano Sterically Stabilized Liposomes (nSSL) to Form Doxil
The need for remote loading
For liposome formulation designed for metastatic tumor treatment, intravenous (i.v.) administration is the only option, and therefore high loading level and its high stability during storage and blood circulation are required. However, due to the combination of needs for nano-liposomes and high dose of doxorubicin (~50 mg/m2) to achieve the sufficiently high loading (to be in the range of intraliposome drug concentration, which is in the range of hundreds of mM] is not an easy task. When the loading is poor, so will be the drug/lipid ratio. This means that either therapeutic levels of drug cannot be reached or therapeutic use of such liposomes will require administering very large amounts of lipids. In addition, when the loading is inefficient there is a great loss of the active agent during the loading and a need to remove unloaded drug. Therefore, the use of liposomes as a vehicle becomes inefficient as well as uneconomical. The major requirements of drug loading into nano-liposomes are (1) to achieve a high level of loading of active agent in the liposome and to make this loading stable during handling and storage, irrespective of the nature of the agent and (2) to fit the release rate of the loaded active agent to specific therapeutic aims of the liposome formulation.
A careful analysis of the currently available loading approaches reveals clearly that the remote loading approach is the best, and in many cases the only, way to achieve the desired intraliposome drug concentration (usually defined as drug to lipid mole ratio) (Barenholz, 2003).
Drug classification
In 1985, we looked for a simple way to classify drugs by their physico-chemical features in a way that will enable the formulator to predict which loading approach to use or, specifically, if a drug is suited for a remote loading approach. At that time, with very little available information, we came up with an oversimplified approach and classified all agents into three categories based on their oil/ buffer and octanol/buffer partition coefficients (/fp). Category I, molecules having high oil/buffer Kp, are considered lipophilic; these molecules do not fit liposomes as their carrier. Category II, molecules having low oil/buffer partition coefficient and medium to high octanol/buffer Kp, are amphipathic. Category III, molecules of very low values in both partition coefficients, by definition, are defined as being water soluble. For some of the molecules, those which are amphipathic weak acids or bases, the classification between group II and III is pH dependent, which determines the level of ionization of the molecule. (Barenholz and Cohen, 1995; Barenholz, 2001,2003)).
Although the use of octanol/water partition coefficient to determine suitability of molecules to reside in a lipid bilayer is controversial, it is well established that it is indicative of agent transmembrane diffusion rate (Stein, 1986), and therefore it is relevant to loading efficiency, loading stability, and the drug release profile.
