WO2001056548A2

WO2001056548A2 - Liposomes composition produced by a dehydration-rehydration process

Info

Publication number: WO2001056548A2
Application number: PCT/GB2001/000391
Authority: WO
Inventors: Brahim Zadi
Original assignee: Lipoxen Technologies Limited
Priority date: 2000-02-04
Filing date: 2001-01-31
Publication date: 2001-08-09
Also published as: DE60123583T2; AU2869901A; JP4786105B2; ATE341310T1; ES2272496T3; US7025988B2; JP2003521508A; US20030138481A1; DE60123583D1; EP1259225B1; WO2001056548A3; EP1259225A2

Abstract

Liposomes containing lipophilic active ingredient are produced by dehydrating a mixture of liposomes, lipophilic active ingredient and sugar, usually followed by a rehydration step to form dehydration-rehydration vesicles. The lipophilic drug is suitably paclitaxel. The sugar is usually sucrose. The liposome forming compounds preferably include cholesterol and phosphatidylcholine, and optionally an anionic lipid. The ratio of sugar:lipid is preferably at least 5:1 w/w. The ratio of lipid:drug is preferably (up to 10):1 w/w.

Description

LIPOSOMES

The present invention relates to liposomes and methods for forming liposomes, as well as their use, in particular in pharmaceutical applications. In particular the invention relates to liposomes containing active ingredients which are lipophilic, of particular utility for formulating paclitaxel and analogues and derivatives thereof.

Paclitaxel is an antimicrotubule agent which has shown promising clinical activity against a wide variety of tumours, including ovarian, breast and lung, as well as melanomas and lymphomas. Derivatives of paclitaxel have also been described having analogous properties. Paclitaxel may be isolated from its natural source (Yew trees), or synthetically produced.

Paclitaxel has a very low solubility in conventional aqueous vehicles. The only preparation approved for clinical use consists of a solution in the vehicle Cremophor EL which is a 50:50 mixture of ethanol and ethoxylated caster oil. The paclitaxel/Cremophor mixture is diluted prior to being infused. One problem is that the diluted mixture is unstable and paclitaxel tends to precipitate, necessitating the use of a filter in line, and resulting in loss of active. Furthermore the organic vehicle produces toxic side effects including an anaphylaxis and pain at the site of injection. Tests on mice indicate a low maximum tolerable dose.

Various ways of improving the dispersability of paclitaxel into aqueous vehicles have been devised. Phosphate derivatives with improved solubility have been made by Ueda, Y. et al in Bioorg. Med. Chem. Lett. 3(1993), 1761-66 and Vyas, D.M. et a/ in Bioorg. Med. Chem. Lett. 3(1993) 1357-60. Paclitaxel has been sialylated to render it more soluble by Takahashi, T. et al in Bioorg. Med. Chem. Lett. 8 (1998) 113-116. It has been complexed with cyclodextrins by Sharma, U.S. et a/ in J. Pharm. Sci. 84(10) (1995) 1123-30. Other prodrugs of Taxol have been described by Mathew, A.E. et al in J. Med. Chem 35(1992) 145-151 and by Deutsch, H. et a/ in J. Med. Chem. 32( 1989) 788-792.

Paclitaxel has also been encapsulated into liposomes. Rahman, A et al in US 5,424,073, US 5,648,090 and (as Cabanes, A. et al) in Int. J. One. 12 (1998) 1035-40 describe a process in which paclitaxel and liposome forming materials are co-dissolved in organic solvent, solvent is evaporated to leave a thin dry film of lipid in drug and liposomes formed by addition of aqueous medium with vortexing and subsequent sonication to form small unilamellar vesicles (SUV's). Rahman et al further describe the incorporation of trehalose into the saline used to hydrate the thin film of the lipid and drug. The liposome suspensions were then stored by being frozen at -80°C (without, apparently, being dried). The sugar-containing frozen liposome mixture is thawed and re-frozen and shown to be storage stable over several months. The mass ratio of sugaπlipid appears to be around 10- 15:1.

Sharma, A. et a/ in Pharm. Res.11 (6) (1994) 889-896 describe encapsulation of paclitaxel (taxol) in liposomes formed from mixtures of phospholipids having an overall anionic charge. The lipids consist of mixtures of phosphatidylcholine and phosphatidylglycerol in various molar ratios. The liposomes are formed by co-dissolving paclitaxel and lipid in organic solvent, forming a thin film of the mixture, re-dissolving in butanol, freeze-dried, then rehydrating in aqueous rehydration medium. Incorporation of negatively charged phospholipid was said to reduce aggregation, and the liposomes were physically stable for more than two and a half months in the hydrated state, provided that the paclitaxel content was no more than 2 mole % (based on total lipid content). The authors indicate that precipitation of the drug in the organic solvent dissolution step caused instability and should be avoided. Bernsdorff C. in J. Biomed. Mater. Res. 46 141-149 (1999) describe various investigations of the effect on liposome bilayers of paclitaxel. The drug appears to affect the fluidity of the bi-layer.

In US-A-5683715 liposomes comprising a taxane comprise, as lipid, essentially only phosphatidylcholine. Examples of phosphatidylcholines are dioleoyl, palmitoyloleoyl or eggphosphatidylcholine. The liposomes may be provided in dehydrated form with protective sugars at the inside and outside surfaces of the bilayer. The general methods described for forming liposomes do not indicate how the taxane is to be incorporated into the liposomes. In the worked examples, paclitaxel is codissolved with lipid in cloroform or methylene chloride. To the mixed solution aqueous buffer is added followed by solvent removal to form multi-lamellar vesicles (MLV). Sonication, to form smaller vesicles resulted in precipitation of paclitaxel. None of the examples freeze dry the liposomal compositions and there is no disclosure of how to incorporate the cryoprotectant.

In US-A-5653998, paclitaxel and other water insoluble compounds are formed into liposomal compositions which contain short chain fatty acids as stabilisers against flocculation. Sugars may be added as cryoprotectants. The formulations are formed by dispersing lipid and taxol directly into water using high energy mixing procedures, followed by filtration and high pressure homogenisation. After the active has apparently been encapsulated, a sugar may be added before freeze drying for cryoprotection. The freeze drying does not change the size of liposomes.

In WO-A-87/01933, amphotericin B, which is a lipophilic active, is formed into liposomal compositions. In one method, the active, sugar and lipid are all codissolved in an alcohol solution, which was dried then hydrated in buffer then freeze dried and rehydrated. In WO-A-99/65465 methods are described for preparing liposomes by a dehydration/rehydration technique. A solution of lipid in an organic solvent is used to coat the inner surface of a suitable vessel and dried to form a thin film. The addition of an aqueous liposome forming liquid results in the formation of multi-lamellar liposomes which are subjected to a size control step to form small unilamellar vesicles (SUV's). Active ingredient may be co- dissolved with the lipid in organic solvent, may be present in the liposome forming liquid or may be mixed with preformed SUV's. The suspension of liposomes which contains active ingredient is subsequently dehydrated, by spray drying or, more usually, by freeze-drying (lyophilization). The dried mixture is subsequently rehydrated in an aqueous rehydration medium to form dehydration/rehydration vesicles (DRV's). The invention in WO-A-9965465 is to incorporate sugar in the suspension of liposomes prior to the dehydration step to increase the level of entrapment of hydrophilic active ingredients.

Merisko-Liversidge, E. ef a/ in Pharm. Res 13(2) (1996) 272-278 describe alternative means for suspending lipophilic drugs in aqueous suspension media. Examples of lipophilic drugs are paclitaxel, camptothecin, etoposide and piposulfan. The active ingredients are wet milled in aqueous suspension containing stabiliser which is a nonionic surfactant. The final particle size was less than 400 nm. According to the present invention there is provided a liposome forming process in which an aqueous suspension containing empty liposomes formed of liposome forming compounds, a lipophilic active ingredient and dissolved sugar is dehydrated.

The product of the dehydration step is a stable product which may be stored for later use. For instance the product may be used as such, for instance administered directly, for instance by inhalation. The dehydrated product may be used after admixture of a pharmaceutically acceptable solid excipient. More usually the dehydrated product is rehydrated before use either as part of a manufacturing step or as part of a medical practitioner's formulation.

In a preferred aspect of the present invention, the dehydrated product is rehydrated in an aqueous rehydration medium to form dehydration/rehydration vesicles (DRV's). In this aspect, the rehydration medium generally comprises a pharmaceutically acceptable aqueous medium, for instance water.

Following the rehydration step, the DRV's may be subjected to one or more size control steps. The DRV's in the rehydration medium may be subjected to physical mixing, sonication or homogenisation. However it is one advantage of the present invention, that the incorporation of sugar in the dehydration step results in the formation of DRV's which have sizes rendering them suitable for administration as a pharmaceutical without further size control steps. Preferably the final product, either of the rehydration step or of subsequent size control steps, has average diameter less than 1000 nm, most preferably substantially all the vesicles having sizes less than 1500 nm. More preferably the average diameter is less than 500 nm. 5 In the present invention, the weight ratio of sugar to liposome forming compounds in the dehydration step is preferably (at least 2):1. Preferably the ratio is (at least 5):1. The ratio may be (20 or more):1 , but is preferably (less than 20): 1. These ratios are on a weight basis, of sugar to total liposome forming compounds. o In the present invention, the ratio of liposome forming compounds:active should be as low as possible. The invention enables ratios of (less than 10): 1 to be achieved. Preferably the ratio is 7.5:(at least

1 ).

In the present invention, the liposomes in the aqueous suspension 5 (before dehydration) are substantially empty of active compounds. They may contain buffer or other liquid excipient. The liposomes are thus formed from liposome forming compounds and aqueous medium in the absence of active ingredient in a previous step or are obtained as such from a commercial source. The liposomes should preferably be small unilamellar 0 vesicles (SUV's) for instance having average diameter less than 500 nm preferably less than 200 nm, for instance in the range 50 to 100 nm. Suitable liposomes are commercially available, or may be produced in a previous liposome forming step using conventional techniques.

Where empty liposomes are used in the process, they are mixed with 5 active ingredient in suitable form and optionally mixed by vortexing or homogenising. Since the active ingredient is lipophilic, it is unlikely to be soluble in the aqueous suspension at the concentration levels required for adequate levels required in the dried product. Suitable ways of dispersing the active ingredient into the aqueous suspension may be as described by o Merisko-Liversidge et al (op.cit).

As an alternative to the wet milling process described by Merisko- Liversidge, the active ingredients may be subjected to a preliminary precipitate or colloid forming step in which it is dissolved in an organic solvent and re-precipitated or formed into a colloid by addition of a non- solvent for the compound, usually water. The precipitate may be collected and resuspended in the aqueous liposome suspension or a precursor aqueous liquid which is added to the remaining components for forming the said aqueous liposome suspension. It may be necessary for the active ingredient dispersion and/or the aqueous liposome suspension to be agitated throughout the process to keep the particles of active ingredient adequately dispersed. A colloidal suspension may be mixed directly with empty liposome suspension.

A preliminary step of forming a suitable precipitate or colloid of active ingredient involves dissolution of the ingredient in an alcohol or an ether. Preferably the alcohol or ether should be sufficiently volatile that it may be removed by evaporation from the precipitate prior to being incorporated into the aqueous liposome suspension. Alternatively the solid and liquid may be separated by centrifugation, with removal of the supernatant, followed by resuspension in water. In some circumstances, however, all of the solvent or a portion of it may remain with the precipitate of active ingredient whereby it remains in the liposome suspension and optionally also in the dehydrated product. The precipitate should have a volume mean diameter measured by a Malvern Mastersizers of about 1 μm to 100μm preferably about 20-50 μm.

Where the organic solvent used to dissolve the active ingredient before precipitate or colloid formation is water-miscible, it may be unnecessary to recover the precipitate as a dry material before mixing with the liposomes suspension. In this case, the organic solvent may remain in the liposome suspension and/or in the dehydrated product, if the organic solvent is involatile. The presence of residual organic solvent may assist the incorporation of the lipophilic ingredient to the product vesicles. Preferably the organic solvent is substantially fully removed during processing, since it may interfere with the properties of the final vesicles. Thus preferably the aqueous suspension which is dehydrated is substantially free of organic solvent. It is preferred to conduct a preliminary active ingredient precipitation step in which the active ingredient is dissolved in a suitable organic solvent, and is then precipitated therefrom by addition of a precipitant. The precipitate is separated from the liquid, preferably by centrifugation or filtration. The supernatant or filtrate is preferably discarded and the solid resuspended in non solvent.

Suitable organic solvents to dissolve the lipophilic active ingredient are alcohols, such as ethanol, or ethers or glycol ethers, such as low molecular weight polyethyleneglycols. The precipitant and resuspension medium are usually aqueous, preferably consisting of water alone.

The sugar used in the dehydration step may be a mono- or di- saccha de. Preferably it is a di-saccharide, for instance trehalose or, most preferably sucrose. The sugar should be fully dissolved in the aqueous suspension. This is generally achieved by forming a precursor solution of the sugar in water, although we have found that some sugars are sufficiently soluble in water that they may be added direct to the liposome suspension as particulate solid.

In the present invention the liposome forming compounds may have an overall charge, which may be positive or, preferably, negative. The use of such compounds may provide improved stability for dehydration rehydration vesicles in aqueous suspension. Anionic lipids are known and include phosphatidylserine and phosphatidylglycerol. It is preferred that the liposome forming compounds have no overall charge, and that each component has no overall charge. Preferably the liposome forming compounds include zwittehonic lipids, usually zwittehonic phospholipids such as phosphatidyl ethanolamine or, most preferably, phosphatidylcholine. The lipids generally comprise di-fatty acyl derivatives (that is esters), but may alternatively include the ether analogues. The fatty acyl groups are selected according to the desired fluidity, by selecting suitable chain lengths and presence, position and conformation of ethylenically unsaturated bonds in the chain. The liposome forming compounds may furthermore comprise cholesterol which may improve stability of the liposomes in the presence of plasma.

The present invention is suitable for encapsulating a range of lipophilic drugs such as those investigated by Me sto-Liversidge et al. op. cit. It is of particular value for encapsulating paclitaxel and derivatives or analogues thereof (taxanes) which, as mentioned above, have particular problems which have proved hard to overcome. Other lipophilic drugs which, it is believed, may be usefully encapsulated in the invention include steroids, platinum-based drugs and non-steroidal anti-inflammatories. The present invention is illustrated in the accompanying examples:

Materials

PC-Hydrogenated soya phosphatidyl choline is obtained from Lipoid

CHOL - Cholesterol is obtained from Sigma

DMPG - Dimyristyoyl phosphatidylglycerol is obtained from Lipoid DSPG - Distearoyl phosphatidylglycerol is obtained from Lipoid

DSPC - Distearoyl phosphatidylcholine is obtained from Lipoid

DSPE-PEG-PEGylated distearoyl phosphatidylethanolamine (PEG

MW = 2000) was a gift from Sequus

EggPC - Egg phosphatidylcholine is obtained from Lipid products. Paclitaxel is the drug used throughout the experiments. The ³H labelled pacitaxel is obtained from Moravek Biochemicals

¹⁴C labelled DOPE (dioleoyl phosphatidylcholine) is obtained from Amersham International.

Methods SUV formation

The liposome forming compounds in the desired ratios (as specified in the tables below) are co-dissolved in chloroform, used to form a thin film on the inner surface of a flask and dried, then water (as liposome forming aqueous medium) is added with vortexing to form multilamellar vesicles (MLV's). The MLV suspension is sonicated and metal particles removed by centrifugation. The SUV's have average diameter in the range 60 to 80 nm. Dehydration and Rehydration Steps

The process is based on the method described by Kirby, C. and Gregoriadis, G. in Biotechnology 2(1984) 979-984. 2 ml of SUV is mixed with sugar in solid form and then active ingredient is added (in the ratios disclosed in the specific examples below) with vortexing, as necessary to avoid any components dropping out of the suspension. The suspension is subsequently cooled to a temperature of around -35°C for 2-3 hours. Subsequently the frozen product is dehydrated at a high vacuum overnight.

The lyophilised cake is rehydrated in 100-300 μl of water (as rehydration liquor) to form DRV's. Samples were taken and the size of the vesicles determined by photon correlation spectroscopy using a Zeta sizer furthermore visual observations were made of the DRV suspension for signs of instability, through flocculation or other problems. Furthermore, the entrapment value for the paclitaxel is determined by observing ³H in the pellet and supernatant after centrifugation.

Example 1 - Optimising sugar level for high lipid:paclitaxel ratio

In this example a solution of paclitaxel in ethanol is made up at a concentration of 20 mg/ml. An amount of the ethanolic paclitaxel solution is added to SUV's formed from the liposome forming compound specified in table 1 , so as to form 1 -4 ml of suspension (depending on the amount of lipid) containing 2 mg paclitaxel and 78 mg total lipid so as to form 4-5 ml suspension. Lipid ratios are molar. An aliquot of solid sucrose is also added using the general method (i.e. before the drug) so as to form a liposome suspension having the sucrose:lipid ratio specified in table 1. The lipid suspension was subjected to lyophilisation and rehydration by the general technique described above. The DRV's were analysed to determine their average size and entrapment values.

The entrapment values were found to be 98 to 99%. The tests were carried out in duplicate and the results are shown in Table 1. Table 1

Example 2 - Varying the lipid amount

In this example the total amount of paclitaxel and of sucrose remain the same. The amount of lipid is varied, but the experiment is otherwise as described in example 1 above. The entrapment levels are again found to be in the range 98 to 99%. The results are shown in table 2 below.

Table 2

The results of examples 1 and 2 indicate that the ratio of sucrose to lipid should be optimised, as should the ratio of lipid to paclitaxel The presence of ethanol in the dehydration stage and/or of paclitaxel appears to increase the ultimate size of the DRV's, where the ratio of lipid drug is too low The presence of sucrose appears not to improve the size to a large degree, although the stability of the DRV's is improved where the lipid.drug ratio is (more than 5):1

Example 3

In this example, the weight ratio of sugar lipid is increased to 10 whilst the pid components are varied In each case the total hpid content is between 38 and 40 mg whilst the paclitaxel level stays as in examples 1 and 2, namely 2 mg Entrapment values for all examples are in the range 98 to 99% Other details are as in example 1 The results are shown in table 3. Table 3

The results in table 3 indicate that cholesterol improves the stability of the DRV's.

Example 4

In this example the significance of the incorporation of anionic lipid is investigated by omission or reduction of the amount of PG or replacement by a PEGylated lipid. In all experiments, the level of lipid is between 38 and 40 mg whilst paclitaxel remains at 2 mg. The same general method as used in Example 1 is followed. Entrapment values were all in the range 98 to 99%. The results are shown in table 4.

Table 4

Table 4 indicates that, for liposomes made by the general method disclosed in Example 1 , the incorporation of anionic lipid appears to confer some stability against flocculation in the product. Reduction of the level of anionic lipid, down to 0.01 parts by mole based on one part of each of PC and cholesterol, still results in liposomes having adequate stability against flocculation. Removing the anionic lipid leads to an unstable suspension, from which the liposomes flocculate. Replacement of the anionic lipid by PEGylated lipid, which might be expected to confer resistance to flocculation, at relatively high levels (Example 4.6) confers some stability. However at the lower level of Example 4.3, the suspension is unstable. Example 5 - Use of PEG300 as Vehicle for Drug In this example, rather than dissolving the paclitaxel in alcohol, it is instead dissolved in an aqueous solution comprising 75% PEG having a molecular weight of 300, in 25% water. Two series of experiments were carried out, one using 100 μl of the 75% PEG solution, and the other using 100 μl of the paclitaxel solution in the 75% PEG solution, in each case added to SUV's in the same manner as the ethanolic paclitaxel solution is added to SUV's in Example 1. Thus the final suspension contains 2 mg paclitaxel and 75 to 80 mg total lipid. The remaining details are as in the general technique.

In each case the entrapment values paclitaxel were found to be around 98%. The lipid mixtures used, the amount of sucrose and the sizes of the produced DRV's are shown in table 5 below.

Table 5

The results in table 5 show clearly the effect on liposomes of the inclusion of sucrose for neutral and for anionic lipids, with and without drug, predissolved in PEG. This is the same general effect as illustrated above for methods in which the drug is dissolved in ethanol.

The results also show that increasing the level of sucrose results in a decrease in the DRV size, again with or without drug. The presence of drug seems to result in larger DRV. Example 6

In this example, a method similar to that used in Example 5 is used. That is the drug is predissolved in a 75% by weight PEG aqueous solution. In this experiment no comparison was carried out with PEG but without drug. In each experiment, the total level of drug used was 2 mg per experiment whilst the amount of lipid was in the range 38 to 40 mg (as compared to 75 to 80 used in Example 5). The lipid mixture is the same as used in Examples 5.1 to 5.4. In each case, the level of entrapment of paclitaxel was around 98 to 99%. The results are shown in Table 6.

Table 6

The results show that, for this system, the increase in the amount of sucrose provides further reduction in the DRV size. Example 7 - In vivo trials

Liposomes made according to the general technique and example 1 , and including sucrose in an amount of 7 times by weight of the weight of the lipid, were made using the lipid mixtures, and ratios and the lipid weightdrug weight ratios indicated in Table 7 below (same compositions as Table 1 ). In order to provide a radio-active tracer in the paclitaxel and lipid, a trace of ³H- labelled paclitaxel and ⁴C labelled dioleoylphosphatidylethanolamine were used respectively. The 1 ml of DRV's were injected i.v. into rats (2 mg paclitaxel and 20, 40 or 78 mg lipid respectively). Samples of blood were taken from the animals after 2 minutes, 10 minutes, 60 to 80 minutes and 120 minutes and the level of ³H and ¹⁴C label detected. The total percentage of the lipid label and the paclitaxel label in plasma were calculated at the respective time intervals.

The results are shown in Table 7. As a comparison, a commercially available paclitaxel in Cremophor composition is used at the same dose level. (The commercially available paclitaxel has a trace of ³H label added). Table 7

The animals were sacrificed after 2^ΛA hours and the level of each label in the kidney(K), liver(L) and spleen(S) determined. The results are shown in Table 8.

Table 8

The results in Table 7 appear to show that there is a dissociation between the lipid and the paclitaxel in the plasma over the period of the experiment. This is likely to be advantageous since it indicates that paclitaxel may be released following administration. The values in Table 8 show that the tissue distribution of the liposomal paclitaxel favours the liver and spleen, as compared to the prior art formulation.

Example 8

Liposomes made as in Example 7 were composed as shown in Table 9. The lipid to drug mass ratios were 20:1. In order to provide a radioactive tracer in the paclitaxel and lipid, traces of ³H-labelled paclitaxel and ^Relabelled dioleoylphosphatidylcholine were used. The liposomes were injected into mice intravenously (0.33 mg paclitaxel and 8 mg lipid for each formulation). Samples of blood were taken at 5, 30, 60 and 180 minutes at which time the animals were killed. Paclitaxel (³H) and lipid (¹⁴C) radioactivities were measured in blood plasma samples, liver, spleen and kidneys and the percentage of the paclitaxel and lipid radiolabels in total plasma and tissues calculated at the respective time intervals. As a comparison, a commercially available paclitaxel (mixed with tracer ³H- labelled drug) Cremophor composition was used in the same dose level.

The results are shown in Figures 1 and 2 and Table 9. Figures 1 and 2 demonstrate that clearance of liposomal paclitaxel (Fig. 1 ) from the plasma is more rapid than that of liposomes (Fig. 2), the implication being that there is a dissociation between the paclitaxel and liposomes in the plasma over the period of the experiment. Fig. 1 also shows that initial (at 5 min) clearance of paclitaxel in Cremophor is more rapid than that of paclitaxel in any of the liposomal formulations shown. Much of the paclitaxel given with liposomes ends up in the liver (L) and spleen (S) whereas lesser amounts of paclitaxel in Cremophor reach these tissues (Table 9). These results show that liposomes can alter the pharmacokinetics of the drug.

Table 9

Example 9 In this experiment, instead of dissolving the paclitaxel in alcohol and using the solution directly to add to SUV's, a preliminary paclitaxel precipitation step is carried out Paclitaxel was solubi sed in absolute ethanol at a concentration of 20 mg/ml. 600 μl of this stock solution (12 mg of paclitaxel) was added to 8 ml of distilled water This resulted in precipitation of paclitaxel, observed by the solution becoming cloudy.

Paclitaxel precipitate was recovered from the suspension by centrifugation The supernatant was removed and the pellet resuspended in 3 ml of distilled water

500 μl of the redispersed paclitaxel (containing 2 mg paclitaxel) was added to a suspension of SUV's, followed by addition of sucrose, dehydration and rehydration following the rest of the general technique described above. In this experiment, the total amount of lipid in each case was 37 to 40 mg, whilst the ratio of sucrose to lipid is shown in Table 9 below, along with the lipid ratios.

Table 10

The results show that, for preprecipitated paclitaxel, the DRV size is smaller for neutral lipid compositions (9.1 and 9.2). Example 9.3, using anionic lipid, also generates desirable low size DRV's. A direct comparison with the neutral lipid mixtures free of PEG is not possible, however, in view of the difference in the level of sucrose. The results of Example 9.4 indicate that, for lipid mixtures with a significant unsaturated acyl group content, the product DRV's have a useful size. Example 10

The general technique indicated in Example 9 for precipitating taxol prior to admixture with SUV's is followed except that, instead of using ethanol, a 75% by volume PEG/water mixture was used to dissolve the taxol (using the same concentration of 20 mg/ml). Furthermore 4 mi of water were used to precipitate 100 μl of solution. After the suspension had been left for one hour, centrifugation is performed to separate the precipitated paclitaxel. The supernatant is removed. The pellet is assessed for its taxol content and found to be around 99% of the originally added taxol. The pellet is redispersed in water, and the suspension mixed with SUV's, and solid sucrose, freeze dried and rehydrated as in the general technique. In each experiment the amount of lipid used for a 2 mg sample of paclitaxel is around 38 mg. The lipid mixtures and the sucrose:lipid ratio are shown in table 10 below.

In this example the size distribution of two repeats were determined with the standard deviation of the size determined for each sample. The results are also shown in the table.

Table 11

The results illustrated in Table 11 show that PEG precipitation with recovery of the paclitaxel from the precipitant, again provides desirable results in terms of the DRV size. Good results are achieved both for neutral lipid mixtures and lipid mixtures having an overall anionic charge.

Example 11

SUV's formed of equimolar soya bean phosphatidylcholine and cholesterol were made using the general technique described above. In order to confirm that liposomes were formed, two sets of experiments were conducted. In the first, the SUV's were mixed with carboxyfluorescein, dehydrated and rehydrated using the general technique described above including sugar in the SUV composition. In parallel, SUV's were mixed with carboxyfluorescein (CF) and paclitaxel, and dehydrated and rehydrated as described above with sugar in the SUV composition. The paclitaxel was in the form used in example 9. The percentage of CF encapsulated, and the Z average diameter were determined. Where CF alone was encapsulated, the rate of encapsulation was about 12%. The Z average diameter was about 100 nm. For the coencapsulated CF and paclitaxel, the rate of encapsulation of CF was about 17%, whilst the Z average diameter was about 171 nm. The behaviour of the carboxyfluorescein in the coentrapped product was similar to that in which CF alone was entrapped. Since CF is well known to become entrapped into SUV's using this technique, it is reasonable to assume that the product of the process of the invention generates small vesicles containing paclitaxel. In a separate test, the SUV's formed of equimolar soya bean PC and cholesterol were mixed with a precipitated paclitaxel, dehydrated and rehydrated using the same general method. The liposomes were recovered and observed under transmission electron microscope. The presence of lipid bilayers could clearly be observed in the DRV's.

Claims

1. A liposome forming process in which an aqueous suspension containing empty liposomes formed of liposome forming compounds, a lipophilic active ingredient and dissolved sugar is dehydrated.

2. A process according to claim 1 in which the dehydrated product is rehydrated in an aqueous rehydration medium to form dehydration/rehydration vesicles (DRV's).

3. A process according to claim 10 in which the rehydration medium consists of water.

4. A process according to claim 1 including the preliminary step of mixing an aqueous suspension of preformed empty liposomes with active ingredient in particulate form.

5. A process according to claim 3 in which the active ingredient is formed into a colloidal suspension in a suspension forming preliminary step before being mixed with the empty liposomes.

6. A process according to claim 3 in which the active ingredient is dissolved in a solvent to form a solution, to this solution a precipitant is added which precipitates the active ingredient, and the precipitate is recovered from supernatant, followed by resuspending the precipitate in a non-solvent.

7. A process according to claim 5 in which the solvent is an alcohol, an ether or a glycol ether, and the precipitant and resuspending vehicle are both water.

8. A process according to claim 6 in which the solvent is selected from ethanol and low molecular weight polyethyleneglycol.

9. A process according to any preceding claim in which the dehydration is carried out by spray drying or, preferably, freeze drying.

10. A process according to any preceding claim in which the sugar is a mono- or di-saccharide, preferably sucrose.

11. A process according to any preceding claim in which the liposome forming materials comprise at least one nonionic or zwittehonic (no overall charge) compound, preferably comprising cholesterol with a phosphatidylcholine.

12. A process according to claim 11 in which the liposome forming materials further comprise an anionic compound.

13. A process according to any preceding claim in which the ratio of sugar: lipid is in the range (more than 5):1 (weight/weight).

14. A process according to any preceding claim in which the ratio of lipid:drug is (less than 10):1 , preferably 7.5:(at least 1 ).

15. A process according to any preceding claim in which the liposomes in the aqueous suspension have average diameter of less than 100 nm.

16. A process according to claim 15 in which the DRV's have a higher average particle size than the average particle size of the liposomes in the aqueous suspension.

17. A process according to claim 16 in which the DRV's have average diameter less than 1000 nm, more preferably less than 500 nm.

18. A process according to any preceding claim in which the active ingredient is a taxane, preferably paclitaxel.