An overview of liposome scaled-up production and quality control

I. INTRODUCTION The history of liposomes as we know them today (vesicles resembling biological membranes) started in the laboratory of Bangham and co-workers 30 years ago, l-4 although at least one medical application of small unilamellar vesicles (SUV) was performed 10 years earlier by Friedman and co-workers.5 They had already demonstrated in 1956 that structures, which today we would call SUV, composed of phosphatidylcholine (PC) injected intravenously into rabbits and other mammals were very efficacious in treating athersclerosis.5·6 It was not until 1990 that the first liposomal product (Ambisome^TM) was approved for clinical use in Europe,7 and in 1995 Doxil^TM8 and Daunoxome^TM were approved in the U.S. Large-scale production of liposomes started in the mid-1980s, late in the short history of liposomes. The reason for the slow development of large-scale production is related to the time it took to resolve technological and quality control issues, including: (1) availability of high-quality lipid raw materials; (2) validated quality control assays; (3) reproducible processes for scaled-up production; (4) fool-proof methods for obtaining sterile (and for injectables, apyrogenic) liposomal products; and (5) long-term chemical and physical stability. All these issues are interrelated; for example, the lack of a complete repertoire of quality control assays slowed down the development of large-scale production and stability assessment,9-12 The cosmetic industry, due to its need for large amounts of liposome preparations,3 was the first to solve these issues. However, the requirements for pharmaceutical dosage forms are much higher, and therefore it took much longer to resolve these issues at the standards required.2.3.9 II. LARGE-SCALE PRODUCTION OF L1POSOMES A. GENERAL ASPECTS The two main obstacles to large-scale production of liposomes are (a) that, despite the very large number of methods available for liposome preparation (see Tables 1 and 2), there is no agreement on the theoretical aspects of how and why liposomes are formed using these different approaches,2·3.9.10 and (b) it is still not clear which of the liposome preparations, if any, are thermodynamically stable assemblies. (See also Chapter 4 by Leung and Lasic.) The energy input required for preparation of liposomes by all methods described in Table 1 is needed either to overcome a kinetic barrier to form a stable assembly or, alternatively, to form a kinetic trap for a metastable assembly. The fact that, at least for phospholipid vesicles, the exact structural parameters of the liposomes are determined to a large extent by the method Table 1 Methods of Vesicle Preparation and Their Liposome Producta b Table 2 Liposome Classification APV Rannie16 (Albertslund, Denmark) for high-pressure homogenization enables the user to up-scale the production from a scale of 35 ml minimal volume per batch using the Rannie 8.30 Minilab one-stage homogenizerl6 having a maximum capacity of 10 llh to a scale of many thousands of liters per hour using higher capacity, two-stage high-pressure homogenizers of APV Rannie or APV Gaulin (Weintraub, Bolotin, Bercovier, and Barenholz, unpublished). The main weak point of these homogenizers is the remote, but existing, possibility that the cooling water of the homogenizer will mix with the liposomal product. Avestin (Ottawa, ON, Canada) offers high-pressure homogenizers under the trade name EmulsiFlex® in a broad volume range, starting with 1 ml minimal volume with a maximum capacity of 20 1Ih in a broad pressure range. Microfluidics Corporation (Newton, MA, U.S.) also makes high-pressure laboratory-and large-scale homogenizers called Microfluidizers® having no moving parts. 17 It is not yet clear, if the scale-up from laboratory-scale to large-scale using Microfluidizers is straightforward. The main advantage of the Avestin homogenizers and of the Microfluidizers is that there is no possibility that during their use the cooling water will mix with the product. The main limitation of the Avestin homogenizers is that, so far, this company does not have high-capacity homogenizers to approach the many-liter scale. The size distribution of vesicles obtained by high-pressure homogenization is, in general, limited to vesicles of <120 nm,9,10.15.16 while preparation of larger and homogeneous unilamellar vesicles, according to our experience, is difficult. Often, in the case of high-pressure homogenization, the product (small vesicles) is contaminated with some giant uni-or oligolamellar liposomes. Another problem is that in the case of lipid mixtures containing a lipid with high aqueous solubility, small mixed micelles are formed and coexist with the vesicles. A post-homogenization extrusion-filtration step may improve the size distribution. Emulsiflex®-C5 (Avestin) is actually a long-awaited combination of homogenizer and extruder. Caution is also needed to minimize oxidation and hydrolysis.