WO1997025026A1

WO1997025026A1 - Drug delivery system

Info

Publication number: WO1997025026A1
Application number: PCT/GB1997/000072
Authority: WO
Inventors: Ruth Duncan; Alexander T. Florence; Ijeoma Florence Uchegbu; Fausto Cociancich
Original assignee: The School Of Pharmacy
Priority date: 1996-01-10
Filing date: 1997-01-10
Publication date: 1997-07-17
Also published as: GB9600471D0; AU1390897A

Abstract

A conjugate of a bioactive compound, for instance an anti-tumour agent, with a polymeric carrier is encapsulated into a vesicle, for instance a non-ionic synthetic surfactant-based vesicle, to alter its bio distribution, prolonging plasma residence time and increasing localisation in tumour tissue. The vesicular system has improved controlled release by cleavage of active drug from the conjugate as compared to the non-encapsulated conjugate.

Description

DRUG DELIVERY SYSTEM

The present invention relates to a vesicular drug delivery system, in which the bioactive compound within the vesicle is covalently bound to a polymeric carrier so as to increase its availability within target cells and increase its circulation time (as compared to unencapsulated drug polymer conjugates) .

It is known also to co-entrap immunological adjuvants with antigenic compounds in liposomes for use as vaccines. Also interleukin-2 has been co-entrapped with antigenic compounds in liposomes to improve their use as vaccines. In O-A-92 04009 antigens or other bioactive molecules may be co-entrapped or covalently linked in liposomal preparations. That reference also describes co-entrapment of interleukin-2 as well as a process for increasing the entrapment rate within small liposomes, said process being useful in combination with or independently of the other aspects of the invention. One of the bioactive molecules said to be useable in O-A-92/04009 is an anti-tumour agent but there is no further explanation as to which aspect of the invention could be applied to anti-tumour agents and no worked examples of such compounds. There is, furthermore, no specific description of the covalent binding of any bioactive molecule with any adjuvant, nor is there a specific suggestion to link active molecules other than antigenic peptides with any other molecule. In Vaccine, 4 (3), 166-172 (1986), Alving et al describe conjugation of synthetic peptide (circumsporozoite peptide) to bovine serum albumin (BSA) followed by entrapment within a liposome and use as a vaccine. The conjugation of peptides to carriers potentiate the immunological response to the peptide antigen. However, in order to retain the immuno-potentiating effect, the peptide must be remain covalently bound to the carrier in vivo. It is known to encapsulate cytotoxic anti-tumour agents into vesicles. Liposomal doxorubicin (adriamycin) formulations are currently undergoing clinical trials.

Free doxorubicin is encapsulated in a liposome in this formulation.

It is also known to reduce leakage of small and/or hydrophobic drug molecules by conjugating them to carrier molecules. For instance S L Law, W Y Lo G W Teh in "The Leakage of Liposome-Encapsulated Adriamycin-Dextran Conjugates" in Drug Development and Industrial Pharmacy 14(1) :143-153 (1988) describe the conjugation of doxorubicin to dextran molecules of various molecular weights. The conjugation is carried out by oxidation of the dextran to form the aldehyde followed by reaction with the amine group of doxorubicin to form an intermediate Schiff base and reduction to form a 2° amine linkage. Such a linkage is not readily cleavable in vivo at useful target sites. It may possibly be cleaved by amine dealkylases for instance found in liver microsomes. In US-A-5037883 drug molecules are conjugated to polymeric carrier molecules via biodegradable linkages in order to prevent glomerular filtration and cellular absorption by simple diffusion. Uptake of the conjugate is consequently restricted to cells capable of pinocytosis, allowing the drug to be targeted to such molecules. The biodegradable linkage is designed so as to be cleavable by lysosomal enzymes, but capable of withstanding- the action of enzymes in the blood stream. The carrier polymer is based on N-(2-hydroxypropyl) -methacrylamide by copolymerisation of peptidyl derivatives of methacrylamide molecules and N-(2-hydroxypropyl)-methacrylamide. The peptide acts as the spacer and is subsequently conjugated to the drug molecule using known chemical synthetic techniques. One conjugate which has been made by the process disclosed in US-A-5037883 is a conjugate of HPMA with doxorubicin, an anti-tumour agent. The conjugate, code named PK1, has been shown to have a beneficial effect in the treatment of experimental neoplasms 1) R Duncan Drug- Polymer Conjugates:Potential for Improved Chemotherapy, Anti-Cancer Drugs, 3:175-210 (1992), 2) L W Seymour, K Ulbrich, P S Steyger, M Brereton, V Subr, J Strohalm, R Duncan Tumour Tropism and Anti-Cancer Efficacy of Polymer based Doxorubicin Prodrugs in the Treatment of Subcutaneous Murine B16F10 Mecanoha, B R J Cancer 70, 636-641 (1994) and the molecule is currently in phase I clinical trials R Duncan, P A Vasey, S B Kay and J Cassidy, Development of PK1, an HPMA Copolymer Doxorubicin Conjugate, The CRC Experience in Proceedings of the EORTC Early Drug Development Meeting 1995, 21-24 June, Corfu Greece (N Pavlidis ED), Ionnina Greece 1960-61, 1995. It would be desirable to increase the plasma residence time for PK1 or any other drug-polymer conjugate. In particular it is found that the conjugate is still excreted to an extent by glomerular filtration and it would be desirable to inhibit this pathway without increasing the polymer chain length. The problem with increasing polymer chain length, which would be expected to extend the plasma half life, would result in a polymer unable to be degraded and excreted thus potentially causing storage diseases.

The present invention solves these problems. New synthetic vesicles according to the invention encapsulate within an aqueous internal space a conjugate comprising a bioactive compound joined by a cleavable covalent linkage to a polymeric carrier.

The synthetic vesicles in the invention are generally sub-micron size, for instance having an average diameter of less than 1 μm, more preferably less than 0.7 μm, for instance in the range 100 to 600 n , especially around 500- 600 nm. Smaller vesicles are preferred since the circulation time tends to be higher and for anti-cancer drugs there is the additional factor that the smaller vesicles can extravasate within the tumour site. The vesicles have a wall formed of at least one bi- layer of amphipathic molecules, optionally including other wall forming components. The amphipathic molecules may be naturally occurring compounds such as lipids, for instance phospholipids, or synthetic analogues of such naturally occurring compounds or may be surfactant compounds capable of forming such bi-layers. The amphipathic molecules may include ionic, for instance cationic and/or anionic compounds or, preferably, are uncharged, for instance being zwitterionic or, most preferably non-ionic. Mixtures of charged and uncharged, for instance non-ionic, molecules may be used, the mixture being selected so as to effect the stability and/or circulation time of the vesicles in use. For instance it is often desirable to incorporate small amount of an anionic compound. Other wall forming components such as cholesterol may provide advantageous properties in terms of stability.

Preferably the vesicles are formed of non-ionic surfactants such as polyalkoxylated long chain carboxylic acids or alcohols. The relative degree of alkoxylation and of length of alkyl chain may be selected so as to provide desirable properties in terms of wall formation and stability. Preferably the surfactant is an ether derivative, that is an alkoxylated long chain alcohol. Most preferably one or more alkyl or dialkyl polygyceryl ethers, for instance as described by Vanlerberghe et al in GB-A-1155712, GB-A-1152713, GB-A-1229234, GB-A-1229235, GB- A-1539625, GB-A-2013609 and FR-A-2358991. Such compounds may be used in combination with ethoxylated alcohols, cholesterol and, optionally also, ethoxylated cholesterol compounds. The use of synthetic amphipathic compounds (as opposed to naturally occurring phospholipids) may increase the stability of the vesicles in vivo by virtue of being less susceptible to attack by lipases which might degrade naturally occurring wall forming compounds and thus destroy the wall. The link between the bioactive molecule and the polymeric carrier is covalent. The linkage must be cleavable in vivo. The linkage can be selected so as to allow the drug to be cleaved from the carrier in the desired target area. The cleavage may be hydrolytic cleavage which takes place at acidic pHs but not at neutral or slightly alkaline pHs. Thus cleavage of the conjugate within the blood stream should be minimal, whilst cleavage would take place when the conjugate is in acidic environments, for instance in certain parts of a cell in the lysozymes, where the pH is around 5.5.

Preferably, however, the linkage is selected so as to be capable of enzymic cleavage, preferably by enzymes present only in selected cells or selected parts of cells. For instance the linkage is cleavable by lysosomal enzymes but not by enzymes present in other parts of the cell or in the circulation. It is particularly desirable for the linkage to include a peptide group which acts as a specific enzyme substrate, the cleavage site generally being a peptide bond (an amide linkage) . One particularly suitable peptide linker group is one which acts as a specific substrate for thiol proteinases, known to be present in lysosomes, as described in US-A-5037883, the disclosure of which is incorporated herein by reference. The polymeric carrier is derived from a polymer molecule having at least one reactive or potentially reactive pendant group and having a molecular weight of at least 15 kD. Preferably the polymer has more than one pendant reactive group per molecule. By reactive group we mean a group on a pendant moiety to which the bioactive molecule or a linker may be attached. In the process of synthesizing the conjugate, that reactive group may be joined to a linker moiety and/or to the bioactive compound after the polymer has been formed or, alternatively, may be already joined to a monomer during polymerisation.

Reactive groups may be, for instance, thiol groups, carboxylic acid groups, amine groups, hydroxyl groups or aldehyde groups, or reactive derivatives of any of these. Preferred reactive groups are hydroxyl groups and carboxylic acid groups, since these can be reacted using conventional peptide synthesis methods, for instance to peptide linker groups and amino acids.

The polymer may be a synthetic polymer or a naturally occurring polymer or derivative thereof. Clearly the polymer should be non-toxic and biocompatible and may, for instance, be biodegradable so as to ensure its excretion from the body after release of the bioactive molecule. If the polymer is biodegradable, then it should be sufficiently resistant to degradation over the period for which it is desired that the conjugate remain intact for controlled release of the bioactive molecule. The polymer must be water-soluble or water-dispersable.

The molecular weight of the polymer is at least 15 kD, more preferably at least 20 kD, for instance 30 kD or more. The maximum molecular weight is generally determined by the water solubility and is preferably 5,000,000 D, most preferably up to 1,000,000 D. The most preferred range for the polymer weight is 30 kD to 200 kD.

Although the polymer may be a naturally occurring compound, such as a carbohydrate or derivative thereof, preferably it is a synthetic polymer, for instance an (alk)acrylate based polymer. A particularly preferred polymer is based on N-(2-hydroxypropyl) methacrylamide, although polymers based on other hydroxy alkyl analogues or on acrylate esters rather than amides may be used. The polymer is preferably substantially as described in US-A-5037883. Other examples of polymeric carriers are polyamino acids, for instance synthetic homopolyamino acids or polypeptides or proteins, polyphosphosines, alginates, dextrans, carbohydrates such as starch and its derivatives, polyethylene glycol having one or two terminal reactive, usually hydroxyl, groups etc.

The bioactive molecule used in the invention may be any molecule which is required to be delivered to intracellular targets. The bioactive compound may be DNA, for instance used in gene therapy or DNA vaccination, when the compound must be delivered into the cell nucleus. The bioactive molecule could be an antigenic compound. In this case, it may be desirable for the polymeric carrier to act as an immuno-potentiating adjuvant after cleavage of the conjugate in vivo. Preferably the bioactive molecule is an anti-tumour drug, for instance dauno ycin, bleomycin, deacetyl colchicine, melanocyte-stimulating hormone (MSH) , desferioxa ine, bestatin, puromycin, melphalan or, preferably, doxorubicin (adriamycin) .

Another class of actives for which the present invention is suited is proteins and peptides. Polymer conjugates of such active compounds have been described previously and any of the conjugates specifically described in the following reference can be used in the present invention:

Duncan, R. , Cable, II.C. , Strohalm, J. and Lopecek, J. Bioscience Reps. 6, 869 877(1987); Flanagan, P.A. , Kopeckova, P., Kopecek, J. and Duncan, R. Biochim.Biophys.Acta. , 993, 83-91(1989); Flanagan, P.A. , Kopeckova, P., Subr, V., Kopececk, J. and Duncan, R. , J. Controlled Release 18, 25-38(1992); O'Hare, K.B., Duncan, R. , Strohalm, J., Kopeckova, P., Kopecek J. and Ulbrich, K. J. Drug Targeting 1, 217-230(1993); and Morgan, S.M., Subr, V., Ulbrich, K. Woodley, J.F. and Duncan, R. Int J. Pharaceutics 128, 99-111(1996).

The present invention has particular additional advantages where the bioactive molecule has a relatively low molecular weight for instance less than 2kD, more preferably less than lkD. For such small molecules the conjugate to a polymeric carrier reduces leakage from the vesicle by permeation through the wall and reduces leakage from a cell into which the vesicle is taken up. The novel vesicles in the present invention encompass vesicles either in aqueous suspension form or in dry form suitable for subsequent suspension, usually lyophilised form. Compositions for use in the treatment of humans or animals are in aqueous suspension form, generally suitable for intravenous administration. Compositions containing the vesicles may be suitable for other modes of administration, such as by inhalation, intramuscular or other means of percutaneous injection, transdermally, through mucosal tissue (e.g. vagina) or by oral administration. The administration of vesicles via the oral route has been described previously in: Cartlidge, S.A. , Duncan, R. , Lloyd, J.B., Kopeckova- Rejmanova, P. and Kopecek, J. J Controlled Release 4, 253- 264(1986); Cartlidge, S.A. , Duncan, R. , Lloyd, J.B., Kopeckova-Rejmanova, P. and Kopecek, J. J Controlled Rel., 4, 265-278(1987); Bridges, J.F., Woodley, J.F., Duncan, R. and Kopecek, J. Int. J. Pharmaceutics, 44, 213-223(1988); Kopeckova, P., Longer, M.A. , Woodley, J.F., Duncan, R. and Kopecek, J. Makromol.Chem.Rapid.Comm 12, 101-106(1991); Pato, J., Mora, M. , Naisbett, B., Woodley, J.F. and Duncan, R. Int, J. Pharm, 104, 227-237(1994); and Morgan, S.M., Subr, V., Ulbrich, K. Woodley, J.F. and Duncan, R. Int J. Pharmaceutics 128, 99-111(1996). Compositions for oral administration may have the vesicles in liquid suspension, formed into tablets or filled into capsules. This invention is of particular utility where the active is a peptide or protein and the route of administration is oral, e.g. for administration of antigens.

The vesicles can be formed by the usual techniques well known in the art. The pre-formed conjugate is contacted with the vesicle wall-forming components under conditions such that the vesicle assembles itself with the conjugate entrapped inside the intravesicular space. The process generally involves formation of multi-lamellar vesicles, dehydration-rehydration of these vesicles and micro fluidisation of the DRVs. A suitable technique is described in WO-A-92/04009, as well as in Kirby and Gregoriadis in Biotechnology, ii, 979-984, 1984. The conjugate can be co-entrapped with stabilisers, such as colloidal stabilisers or suspending agents eg polymers such as polyvinyl pyrrolidone, polyethylene glycol, carbohydrates (such as alginate) or sugars (such as lactose) . The advantages of the present invention are that the system provides improved circulation time and consequently improved location at the target tissue for the bioactive compound as compared to the unencapsulated conjugate. As compared to encapsulated free drug, the conjugation to the polymeric carrier provides improved increased period of availability of drug inside a target cell in view of reduced rate of permeation of conjugate through the cell wall as compared to the smaller drug molecule and enhanced controlled release of free drug into the target tissue. In addition, the conjugatin to the carrier provides stability and reduced leakage of drug from the vesicles and improved rate of encapsulation by virtue of hydrophobic drugs being made more water soluble by conjugation to a more hydrophilic and water soluble polymer carrier. As compared to the polymeric carrier used in the Law et al paper (op citi , the use of a cleavable linker group allows the release of drug from the conjugate in a controlled manner at the target site. This invention can provide two step delivery following any route of administration. Thus there are opportunities for targeting and 2 step controlled release. This could be particularly interesting for controlled release of a bioactive in the gastrointestinal tract, for example an antibiotic (towards 24h formulation rather than 3 x daily) . Alternatively for protection of a protein or peptide drug in the gastrointestinal tract (GI) . This can be useful in oral immunisation (using a peptide antigen) and also for peptide and protein delivery. The system can:

1) stabilise against degradation 2) facilitate local delivery in the correct region of the GI tract 3) both the vesicle and the polymer conjugate have been shown to improve transport across the GI.

This process is endocytosis across the tissue

(transcytosis) . The present invention is illustrated further in the following examples and drawings, in which:

Figure 1 shows the chemical structure of an embodiment of a polymer-drug conjugate, PKl;

Figure 2 is a flow diagram showing the steps in preparing vesicles for storage stability tests in Example 3;

Figures 3 to 10 show the results of storage tests carried out in Example 3 indicating the change of vesicle size and amount of PKl entrapped upon storage under different conditions;

Figures 11 and 12 show the entrapment rates for tests carried out in Example 3 showing the effect of co- entrapping an additional polymer, or using polymers as suspending agents; Figure 13 shows the effect of changing pH on the release of PKl from vesicles of Example 3;

Figures 14 and 15 show the release of doxorubicin from vesicles in the presence of buffers, plasma and tritosomes (lysosomal enzymes) . Materials

C₁₆E0₅, (ie 5-mole ethoxylated C₁₆-alkanol) C₁₈E0₅ (ie 5- mole ethoxylated C₁₈-alkanol) , sorbitan monostearate (Span

60) and cholesterol and dicetyl phosphate, (DCP) , MTT, daunomycin, Triton X-100 and sodium hydroxide pellets were all from Sigma Chemical Company, UK. All organic solvents were from BDH, (UK) and were of analytical grade. PKl (FCE

28068) a doxorubicin-N-(2-hydroxypropyl) methacrylamide copolymer conjugate incorporating a glycine-phenylalanine- leucine-glycine peptidyl spacer made as described in US-A-5037881 and illustrated in Figure 1 was from

Farmitalia Carlo Erba (now Pharmacia) , Milan. Solulan C24

(a poly-24-oxyethylene cholesteryl ether) was from UK supplier, Ellis & Everard (Anstead International) (US main producer A erchol) . C₁₆G₂ (C₁₆-alkyl diglyceryl ether) was a gift from L'Oreal, France.

EXAMPLES 1/2 - Investigation of various surfactant combinations

1. Preparation of vesicles

Vesicles were prepared by the dehydration rehydration vesicle (DRV) technique (Kirby and Gregoriadis, 1984 op. cit.) and by various modifications therefrom. In general the ratio of the non-ionic surfactant to cholesterol was maintained at 1:1, while the level of the polyethoxylated cholesterol compound Solulan C24 was altered. The surfactants and levels are specified in Tables 1 and 3 below. As high levels of Solulan C24 are potentially haemotoxic, the haematotoxicity of a few of the formulations was studied. Variations in the method of vesicle preparation were also assessed. The vesicles were sized as described in Example 3 below. The entrapment rate for the vesicles was determined as described. The results are shown in Tables 2 and 3 below.

Method a

A thin film of surfactant/lipid films (150 μ oles total lipid/surfactant) in the proportions indicated in Table 1 was hydrated with 2.5 mL water and sonicated for 4 mins on an MSE PG100 150 W probe sonicator. Samples were centrifuged (Haraeus Varifuge 3.0RS) at 3,700 g for 10 min to remove titanium particles arising from probe sonication. To the ensuing supernatant of empty non-ionic vesicles (encapsulating water) was added l mL (10 mg mL¹) . The mixture was shaken for 16 h, ImL aliquots were frozen dried cake which was resuspended with vortexing. Some of the resuspended freeze dried material was filtered (0.45 μm and 0.22 μm filters) and the filtered and unfiltered suspension then treated by separating unentrapped material by ultracentrifugation for 2 X 1.5 h at 145,000 g. The pellet was sized as discussed in Example 3 and assayed as described below. The results are shown in Table 2. Method b

Sonicated non-ionic vesicles were added to a solution of PKl and immediately freeze dried thus omitting the 16 h incubation step of method a. Filtration subsequent to freeze drying and the separation procedures (ultracentrifugation) were as described above. The surfactant lipid proportions are specified in Table 3. The initial level of lipid in the method was 75 μmoles whilst the level of PKl was 20 mg. Assay of vesicles for PKl

Vesicle suspensions were diluted in 2 parts isopropanol and centrifuged for 10 mins (3,700 g) to rupture the vesicle wall. Absorbance of the supernatant was read at 478 nm absorbance being proportional to doxorubicin concentration in the supernatant. The supernatant from the ultracentrifugation process described above was also assayed in a similar manner. To calculate % entrapment: a%- _«e_«n•t•-r-a=_«p_«m_«e_«n₊t- -- l_ιnO_nO_vx A^mou^nt _oof_f D_Dr_ru_u ^g _g i _in_n P_Pe_elllϊe_ett +

Amount of Drug in Supernatant 2. Hae otoxicity experiments

Non-ionic vesicles were prepared as described above except that after the preparation of the empty vesicles (encapsulating water) , the suspension was then frozen in liquid nitrogen and freeze dried. The sample was not filtered after rehydration. The assessment of haemotoxicity was as previously described (Duncan et al, 1991 "Soluble polymeric drug carriers: haemocompatibility" in Progress in Membrane Biotechnology ed. Gomez-Fernanez, Chapman and Packer 253-265) . Briefly blood was obtained by cardiac puncture of Sprague Dawley rats and immediately erythrocytes were separated from other blood products by centrifugation at 1,000 g for 10 min. Erythrocytes were washed three times in PBS, pelleting each time. Pelleted erythrocytes were weighed and a 2% w/w solution of erythrocytes in PBS prepared. To 1 ml of this solution was added 1 ml of the niosome suspension (encapsulating water) at concentration to give the level of lipid (including surfactant) specified in the table. 1 L PBS was used in control incubations and a final concentration of 1% triton- X 100 was used to achieve 100% haemolysis. The mixture was incubated for 5 h at 37°C. Released haemoglobin was separated from erythrocytes by centrifugation at 4,000 g for 10 min. The supernatant (containing haemoglobin) was diluted with an equal volume of isopropanol and absorbance read at 550 nm, relative to the similarly treated supernatant from the control (PBS) incubation. The results are shown in Table 4.

Table 1: Vesicle Formulations Studied

TABLE 2

EXAMPLE MEAN SIZE % ENCAPSULATION nm

1.1 215 28

1.1F 200 23

1.2 145 5

1.2F 150 6

1.3 205 2

1.3F 150 4

1.4 530 17

1.5 800 4

1.5F 175 3

1.6 370 64

1.6F 155 45

1.7 270 40

1.7F 250 31

1.8 390 3

1.8F 370 2

1.9 285 30

1.10 570 7

1.10F 140 5

F after the example refers to filtered material (see method)

in in TABLE 4 - HAEMATOTOXICITY

FORMULATION NIOSOME LIPID ABSORBANCE RELATIVE TO

OF EXAMPLE LEVEL μM PBS INCUBATION (λ=550 nm)

1.6 0.113 0.0

0.225 0.0

1.125 0.0

4.5 0.3

1.9 0.113 0.0

0.225 0.01

1.125 0.07

4.5 0.13

1.7 0.75 0.16

1.5 0.26

7.5 0.26

1.9 0.75 0.09

1.5 0.20

7.5 0.27

The control haemolytic agent 1% Triton X400 gives an absorbance level of 0.29. This is equal to haemolysis.

Discussion

These results seem to show that optimum entrapment efficiency is obtained where the level of polyethoxylated cholesterol is relatively low (in these examples 9%) . There is some correlation of entrapment efficiency with size, with higher levels of polyethoxylated cholesterol giving low vesicle size. The alkyl polyglyceryl surfactant containing vesicles have high entrapment rates as compared to vesicles based on the structurally similar polyethoxylated alcohols.

Filtration of the vesicles reduces the mean size by removing larger vesicles. The reported entrapment percentage is also reduced to an extent.

The results of the haematotoxicity tests show that the alkyl polyglyceryl based vesicles have lower haematotoxicity at low levels than the sorbitan monostearate based vesicles, where the level of polyethoxylated cholesterol is relatively low. However for higher levels of polyethoxylated cholesterol the alkyl polyglyceryl based vesicles seem to have higher haematotoxicity than the sorbitan monostearate based vesicles.

The alkyl polyglyceryl compounds also give low polydispersity values which is desirable as it shows a narrow size distribution, giving better control of drug delivery.

By incubating preformed niosomes with PKl solution, PKl solutions do not come into contact with high temperatures which cause PKl to precipitate in high concentration. It has also been found that omitting the 16h incubation step does not affect the % entrapment. Loading PKl onto niosomes containing PBS did not enhance the % entrapment when freeze-drying was omitted.

EXAMPLE 3

Method of preparing vesicles

A total of 150 μmoles of the compounds Span 60: cholesterol: Solulan C24 (a polyoxyethylene-24-cholesteryl ether) were weighed in the molar proportions 45: 45: 10, dissolved in chloroform and a thin film produced by rotary evaporation under reduced pressure at a temperature of 45°C. 5 ml of a 3 mg ml^" solution PK, in water was added to the dried film and the mixture shaken for 1 h at 60°C. A homogenous dispersion is formed which was sonicated (Soniprep 150 MSE at 2/3 of its maximum output) for 5 lm periods with 30s in between for cooling. The production of dehydration rehydration vesicles was carried out after the dispersion was allowed to cool slowly to room temperature. 1 mL volumes were flash frozen in liquid nitrogen and the lyophilized overnight (Edwards Modulyo freeze drier) . Rehydration of the lyophilized preparation was carried out the following day by the addition of 1 mL water in 0.2 mL fractions at room temperature with rapid vortexing. Separation of unentrapped material was by ultracentrifugation at 145,000 g for two 1.5 h cycles with the supernatant being replaced by fresh water after each cycle. In order to improve the entrapment of PKl in the vesicles some samples were prepared rehydrating and purifying the niosomes with an aqueous polymer solution of lactose, polyethylene glycol having an average molecular weight of 8000 (PEG 8000) polyvinylpyrrolidone (povidone) or alginate. These solutions were prepared equimolar with the concentration of PKl entrapped in the vesicles (5.03 x 10^*5 mmol/ml) except for the alginate solution (1.1 x 10^*5 mmol/ml) because of the low solubility of alginate in water. Two different methods based on the method of this example were compared: in the first one the vesicles were rehydrated with 1 ml of the aqueous polymer solution while in the second one after the rehydration with 1 ml of water 1 ml of the aqueous polymer solution in double strength was added.

The results of the various tests are shown in the figures indicated following the description of the method. In the figures the vesicles are referred to as 'niosomes' which is a trade mark of L'Oreal. Entrapment Efficiency

0.5 ml of the vesicle suspension was added to 1 ml propan-2-ol to disrupt the vesicles. Vesicles were destroyed once propan-2ol was added but insoluble lipid or surfactant material is removed by spinning. The amount of PKl entrapped (ie which was thereby released) was determined spectrophotometrically (UV-1601 Spectrophotometer, Shimadzu) at a wavelength of 478.0 nm. Figure 11 compares entrapment rate for the systems with co- entrapped polymer whilst the rates for the second method in which polymer solution is mixed after rehydration are shown in Figure 12.

Measurement of vesicle size bv photon correlation spectroscopy

A Malvern photon correlation spectrometer (Malvern Instrument 48 channel, 7023 single clipped photon correlation spectrometer) was used for determining the average diameter of the vesicles. 10 μl of the suspension was diluted with 2.5 ml distilled water (double filtered, syringe filr 0.45 mic.) into cleaned absorbance cells. The decay curve of the diluted suspension was measured using He/Cd 10 mW laser λ=441.6 nm at 25°C and scattering angle 0=90°C. The spectrometer can also report the polydispersity, that is size distribution of samples, as reported in Table 3 of Example 2 above.

The mean sizes of vesicles as made, with and without co-entrapped polymer are shown in Table 5 below.

Table 5 MEANSIZE(nm) OFTHE NIOSOMES PREPAREDWITH AQUEOUS SOLUTION

REHYDRATION WITH THE REHYDRATION WITH POLYMER SOLUTION WATER THEN ADD THE POLYMER SOLUTION

WATER 483.2 ± 63.3 483.2 ± 63.3

ALGINATE 701.4 ± 150.7 -

POVIDONE 342.3 ± 43.7 538.0 ± 100.4

PEG 8000 365.2 ± 46.6 514.2 ± 144.8

LACTOSE 387.6 ± 68.3 -

Evaluation of the pharmaceutical stability of the vesicles After freeze-drying the batch of vesicles was divided in four different types, as shown in Figure 2, to store. The temperature of storage was 4 and 25°C for the washed and unwashed aqueous suspensions and -40, 4 and 25°C for the vesicles stored as a lyophilised suspension. Over a period of one month of storage each different type of vesicle was analysed in order to determine the entrapment of PKl ("pharmaceutical stability") and the mean size. The results are shown in Figures 3 to 10. In vitro release studies at different PH

The vesicles were prepared as described above but using PBS (pH 7.4) instead of water. After the purification process the pH of the half samples was adjusted to pH 5.5 with HCl IN. The release study was performed at 37°C and samples were taken within 24 hours, ultracentrifuged (146,600 ref for 90 min) and the amount of PKl in the supernatant was determined spectrophometrically as described above. The results are shown in Figure 13. Incubation in vitro with Tritoso es

The vesicles were prepared in water as described above and they were resuspended in 1 ml citrate-phosphate buffer (pH 5.5) after the purification process. 400 μl suspension was then incubated with 400 μl Tritosomes (a lysosomal enzyme mixture) in the presence of 100 μl EDTA solution (10 mM in citrate-phosphate buffer) and of 100 μl reduced glutathione (50 mM in citrate-phosphate buffer) . In order to prevent lysis of vesicle membrane in the presence of Triton X-100, that is usually add to the citrate-phosphate buffer to destroy Tritoso e membrane, the enzyme membrane was destroyed by thermal shock. As a control the non-ionic vesicles were also incubated with citrate-phosphate buffer instead of Tritosomes. The incubation was performed at 37°C in a water bath within a period of 72 hours. Because of the impossibility to separate the vesicles and the supernatant in the presence of enzymes (ultracentrifugation, ultrafiltration, gel chromatography) only the doxorubicin release was analyzed. 100 μl samples were taken off at each time point refrigerated and extracted as described below before the HPLC analysis. The results are shown in Figure 14. Incubation in vitro with plasma

To carry out in this experiment the same procedure of the incubation with Tritosomes was followed, but all the solutions were prepared in PBS. The plasma was used instead of Tritosomes and it was obtained by separation of the rat blood collected through cardiac puncture. The results are shown in Figure 15. Extraction method of the samples All the procedure [wedge thesis] was performed in polypropylene centrifuge tube in darkness and at 4°C in order to prevent doxorubicin instability. To the sample 100 μl daunomycin (100 ng, 1 μg/ml stock solution) was added as an internal standard and 800 μl water to make up to 1 ml. Then 100 μl ammonium formate buffer (IM, pH 8.5), 5 ml of chloroform:propan-2-ol (4:1) mixture were added and the mixture was vortex. After centrifugation (1000 ref, 30 min) the aqueous layer was carefully discarded by aspiration and the organic phase evaporated using a nitrogen apparatus. The dried sample was resuspended in 100 μl methanol previous the HPLC analysis. In order to detect the doxorubicin attached to HPMA (PKl) the sample was subjected to acid hydrolysis thereby cleaving the glycosic bond of the doxorubicin molecule to liberate aglycone that could be extracted and analysed following the same method used for doxorubicin. After addition of the internal standard 1 ml HCl (2M) was added to the sample and heated at 80°C for 25 min. After cooling 1.5 ml ammonium formate buffer (IM, pH 8.5) and 1 ml NaOH (2M) were added. Then the extraction procedure was followed. HPLC analysis

Analysis was performed using a C-18 μbondapak reverse phase HPLC column, with the mobil phase (29% propan-2-ol in water, adjusted to pH 3.2 with orthophosphoric acid, filtered and degassed) driven by a Constametric III pump (Milton Roy) at a flow rate of 1 ml/min. Fluorescence was measured with a fluoro onitor II (Milton Roy) , filtered with interference filters at 480 nm for excitation and 560 nm for emission.

Discussion

The above results show that the purification techniques investigated have no significant effect on changing the vesicle size after no storage. It appears that the various purification techniques have some effect on storage stability although in general all the selected storage conditions seem to retain the same average size thus indicating that they do not aggregate (or become disrupted) on storage. The entrapment values show that for all forms of storage there is some leakage of active material from the liposome during the first few days of storage. This initial loss of entrapped active ingredient seems to worse, the more steps in the purification procedure. The graphs showing entrapment quantity for the various suspensions start at the same point at day 0, since only one batch was made, assayed and portions stored in the modes specified. The amount of active material entrapped in the "lyophilised washed suspension" is reported as entrapment quantity at day 0 for all the samples.

Although the results shown in Figure 12 show that for one method of co-entrapment of PKl and another polymer show that the other polymer has little effect on the entrapment rate, Figure 12 shows that where the method is improved, both polyethylene glycol and polyvinylpyrrolidone can improve the entrapment rate whilst having no adverse effect on leakage of PKl from the niosomes. Table 5 shows that the first method with polymer, apart from alginate, has the effect of making the liposomes smaller whilst the second has no significant effect on size.

Figure 13, namely release of PKl from the niosomes at different pH, shows pH does not drastically alter the availability of PKl from niosomes. This makes the fact that more DOX is available in the presence of lysosymes (pH 5.5) than in the presence of plasma (7.4) an important result and emphasizes the importance of enzymatic cleavage. Figure 14 shows that in the presence of tritosomes (containing lysosomal enzymes) DOX is released from the niosomes at a greater rate than in the presence of citrate- phosphate buffer.

The results shown in Figure 15 show that the niosomes are stable in the presence of plasma, that is the vesicles are not disrupted so that no doxorubicin is released.

Claims

1. Synthetic vesicles encapsulating a conjugate within an aqueous internal space which comprises a bioactive compound joined by a cleavable covalent linkage to a polymeric carrier.

2. Vesicles according to claim 1 formed from synthetic surfactants, preferably primarily non-ionic surfactants, most preferably polyalkoxylated alcohols, most preferably including alkyl or dialkyl polyglyceryl compounds and/or polyethoxylated alcohols, in any of which the alkyl group contains 12 or more carbon atoms, preferably 15-24 carbon atoms.

3. Vesicles according to claim 1 or claim 2 which include cholesterol and/or a cholesterol derivative.

4. Vesicles according to any preceding claim in which the conjugate is stable at one pH value and is cleavable at a different pH value found inside cells.

5. Vesicles according to any of claims 1-3 in which the bioactive compound is bound to the polymeric carrier via a linkage which is enzymically cleavable, preferably comprising a polypeptide moiety.

6. Vesicles according to claim 5 in which the bond which is cleavable is a peptide bond.

7. Vesicles according to claim 5 or 6 in which the linkage is susceptible to hydrolysis by lysosymal enzymes but withstands enzymic hydrolysis by enzymes found in the blood stream.

8. Vesicles according to any preceding claim in which one molecule of the polymeric carrier is conjugated to more than one molecules of bioactive compound, preferably an average of 3 or more.

9. Vesicles according to any preceding claim in which the polymer comprises recurring units derived from N-(2- hydroxypropyl) methacrylamide.

10. Vesicles according to any preceding claim in which the bioactive compound is an anti-tumour agent.

11. Vesicles according to any of claims 1 to 9 in which the bioactive compound is an oligo nucleotide.

12. Vesicles according to any of claims 1 to 9 in which the bioactive compound is a polypeptide or protein, preferably an antigen.

13. Vesicles according to any preceding claim in which the bioactive compound has a molecular weight of less than 2 kD, preferably less than 1 kD.

14. A pharmaceutical composition comprising vesicles according to any preceding claim and a pharmaceutically acceptable excipient.

15. A composition according to claim 14 which is an aqueous suspension, preferably suitable for percutaneous injection of a human or animal.

16. A composition according to claim 14 which is for oral administration preferably being in the form of a tablet comprising vesicles or capsules containing vesicles.