WO2006048639A1 - Hydrogel delivery vehicle - Google Patents

Abstract

Cyclodextrins, particularly hydroxypropyl-β-cyclodextrin, are found to improve the uptake of poorly soluble hydrophobic active agents into hydrogel delivery vehicles, such as pessaries, suppositories etc; and to improve the delivery therefrom. The water-swellable hydrogel is generally formed of a cross-linked polyurethane formed of a diol, triol and isocyanate. The cyclodextrin may be a α-, β- or Ϝ- cyclodextrin having 6, 7 or 8 glucose units respectively. This may be substituted or unsubstituted.

Description

HYDROGEL DELIVERY VEHICLE

The present invention relates to a pharmaceutical delivery vehicle comprising a water- swellable hydrogel body for sustained release of pharmaceutically active agents. In particular, the present invention is directed to the problem of successfully loading pharmaceutically active agents which are hydrophobic into the hydrogel body, and also to enabling successful delivery of such hydrophobic active agents from the hydrogel body into aqueous environments existing within a human or animal patient.

Solid hydrogel delivery deviόes, such as pessaries or suppositories are well known in the art and have been available commercially for the delivery of pharmaceutically active agents such as prostaglandins to assist in child birth. Such hydrogel bodies are water- swellable, often up to several times their original volume, whilst at the same time retaining their integrity. They are therefore different from essentially liquid or fluid forms such as liquids, pastes, creams etc. Suitable cross-linked polyurethane hydrogels are disclosed in patent specifications EPOO 16652 and EPOO 16654 (NRDC) and patent specification WO/00/32171. Linear uncrossed-linked hydrogel bodies are also described in patent specification WO/2004/029125. These patent specifications describe the methods of production of the hydrogels, methods of determining the percentage swelling etc.

Typically, pharmaceutically active agents are loaded into the solid hydrogel body by swelling the hydrogel in an aqueous solution of the drug, thereby allowing the drug to be taken up into the hydrogel, followed by drying the hydrogel body. When a hydrogel body (typically as a pessary, suppository or buccal insert) is inserted into a patient, the hydrogel swells within the aqueous environment and allows the pharmaceutically active agent to be released in a sustained, controlled and reproducible way. Thus, essentially constant release of active agent may be achieved over a certain time period. This avoids problems of efficacy and toxicity in the patient. However, such hydrogel delivery vehicles are at present essentially only suitable for water-soluble active agents. Hydrophobic pharmaceutically active agents are difficult to load into the hydrogel body and are difficult to release effectively from the hydrogel into the aqueous environment in the patient due to their low water-solubility. It is possible in some instances to choose a solvent system (typically a mixture of solvent systems such as water and alcohol etc.) in which the hydrophobic active agent is soluble and which allows the active agent to be loaded into the hydrogel body. However, the problem of successfully releasing the hydrophobic active agent from the hydrogel delivery device into the aqueous environment within the patient remains unsolved.

An object of the present invention is therefore to provide a suitable formulation which allows hydrophobic pharmaceutically active agents to be employed in hydrogel delivery vehicles for the treatment of patients.

Many solubilising agents for use in pharmaceutical compositions are available in the literature, such as polyethyleneglycols, surfactants, cyclodextrins etc. The present invention has found cyclodextrins to be particularly valuable.

Cyclodextrins

Cyclodextrins are crystalline, non-hygroscopic, cyclic oligosaccharides derived from starch. They are "bucket-like" or "conelike" toroid molecules, with a rigid structure and a central cavity, the size of which varies according to the cyclodextrin type. The internal surface of the cavity is hydrophobic and the outside of the torus is hydrophilic; this is due to the arrangement of hydroxyl groups within the molecule. This arrangement permits the cyclodextrin molecule to accommodate a guest molecule within the cavity, forming an inclusion complex. The advantage of cyclodextrins is that they are non-toxic and non-irritant materials (Rowe, R.C., Sheskey, PJ. and Weller, P.J. (Eds.), Cyclodextrins, In: "Handbook of Pharmaceutical Excipients", 4^th Ed. Pp. 186-190, 2003).

The three main types of cyclodextrins are α-, β- and γ-cyclodextrin (see Figure 1) which are made of six, seven and eight glucose units, respectively. β-Cyclodextrin is the most widely used but it is the least soluble, while γ-cyclodextrin has the largest cavity and highest solubility among the three cyclodextrin types (see Table 1).

Tablel. Types of Cyclodextrins (Rowe, et al., 2003).

Cyclodextrins may be used to form inclusion complexes with a variety of drug

molecules, where the guest molecule associates with the cyclodextrin so that the hydrophobic

portion of the guest molecule interacts with the hydrophobic cavity of the cyclodextrin. This

interaction is an equilibrium reaction. Cyclodextrin complexation results primarily in

improvements to dissolution and bioavailability owing to enhanced solubility and improved

chemical and physical stability.

Cyclodextrin inclusion complexes have also been used to mask the unpleasant taste of

active materials and to convert a liquid substance into a solid material and to improve quality

and functional properties of pharmaceuticals, foods and agricultural products (Makela, M.J.

Korpela, T.K., Puisto, J. and Laakso, S.V., Nonchromatographic Cyclodextrin Assay:

Evaluation of Sensitivity, Specificity and Conversion Mixture Application, J. Agric. Food

Chem., 36, 83-88, 1988). Hydroxypropyl-β-Cvclodextrin

Because β-cyclodextrin has low water solubility and cannot be used parentally, derivatives of β-cyclodextrin have been developed. Hydroxypropyl-β-cyclodextrin (HPBCD) is a partially substituted poly(hydroxypropyl) ether of β-cyclodextrin.

The basic closed circular structure of β-cyclodextrin is maintained in HPBCD. The glycosidic oxygen forming the bond between the adjacent glucose monomers and the hydrogen atoms lining the cavity of the cyclodextrin impart an electron density and hydrophobic character to the cavity. Organic compounds interact with the walls of the cavity to form inclusion complexes. The hydroxyl groups and the hydroxypropyl groups are on the exterior of the molecule and interact with water to provide the increased aqueous solubility of the HPBCD and the complexes made with the HPBCD.

Degree of Substitution

The hydroxypropyl groups are randomly substituted onto the hydroxyl groups of the cyclodextrin and the amount of substitution is reported as average degree of substitution or number of hydroxypropyl groups per cyclodextrin and is the preferred manner of describing the substitution.

Degree of substitution can have an effect on the binding of a guest molecule to the HPBCD molecule. At low degrees of substitution, binding is very similar to that of the unmodified β-cyclodextrin. Increasing substitution can lead to weakened binding due to steric hmderance. The effect is dependent upon the particular guest molecule and it is also possible to obtain increased binding due to an increase in surface area to which the guest molecule can bind. With most guest molecules these differences in binding with degree of substitution are small if detectable. Unlike β-cyclodextrin , HPBCD is not nephrotoxic and it has been used in parental formulations. It has a much higher water solubility of 1 in 2 parts than β-cyclodextrin at 25°C, which is 1 in 50 parts. The United States FDA has approved some pharmaceutical formulations that contain HPBCD, such as:

Itraconazole 10mg/ml (Sporanox TM l.V and Sporanox . TM liquid) (antifungal)

• Tolnaftate (antifungal)

• Mitomycin for injection (Mitozytrex ) (for cancer).

HPBCD used herein is a white crystalline powder, with a molecular weight of about 1380 and a degree of substitution of 0.6.

Pharmaceutical compositions comprising cyclodextrins are known, for example, from patent specifications US 4,596,795, US 4,727,064, US 6,407,079, US 2,002,0019369 and US 2,002,0192243. None of these prior art documents discloses the use of cyclodextrins in hydrogel delivery devices.

Invention

Generally speaking, the present invention is based on the surprising discovery that cyclodextrins may not only be effective in assisting the loading of hydrophobic pharmaceutically active agents into the hydrogel body, but surprisingly may assist the delivery of the hydrophobic active agent into aqueous biological systems where they can exert an effective pharmaceutical action. One or more mechanisms may be present.

More specifically, the present invention provides a pharmaceutical delivery vehicle,

which comprises: a water-swellable hydrogel body containing a pharmaceutically active agent together with a pharmaceutically acceptable cyclodextrin.

The cyclodextrin may be any pharmaceutically acceptable non-toxic cyclodextrin and may be an α-, β- or γ-cyclodextrin which is substituted or unsubstituted. Sulfobutylether β cyclodextrin may be used. Clearly, cyclodextrins which exhibit nephrotoxicity should be avoided. The hydroxyl groups of the cyclodextrin may be substituted by a variety of hydrophilic substituents such as hydroxyalkyl or poly(hydroxyalkyl) groups (e.g. having a formula H(O-C₂H₆) _n) - where the alkyl group is C₁-C₆ alkyl or sodium salts of sulfobutyl ether groups (0-C₄H₈SO₃Na)_n. HPBCD is preferred. However, the present invention envisages the use of any suitable non-toxic cyclodextrin or substituted cyclodextrin molecule.

Generally, the water-solubility of the cyclodextrin (which term includes substituted cyclodextrins) is in the range one part cyclodextrin in 0.1 to 100 parts water at 25°C, typically one part cyclodextrin in 1 to 10 parts water.

The cyclodextrin may operate in a number of ways. Firstly, the presence of the cyclodextrin in the solution used to load the active into the delivery vehicle may improve the solubility of the active agent therein; thus potentially allowing greater amounts of active to be loaded into the delivery vehicle. Secondly, the presence of the cyclodextrin may enhance transport of the active agent further into the delivery vehicle, so it does not merely remain near the surface thereof. This tends to prolong long term delivery of the active agent from the delivery vehicle and minimises the burst effect (a rapid immediate release of active agent). Thirdly, the cyclodextrin may improve the delivery of the active agent from the delivery device into the patient by enhancing the water-solubility of the active agent. This may be important for release in vivo into body cavities (e.g. the rectum or vagina) where the water content is relatively low. One or more of these mechanisms may be operative.

A further factor is that the presence of large amounts of the cyclodextrin may tend to inhibit uptake of active agent, since the space within the matrix of the delivery vehicle available to take up additional molecules is limited. Thus, use of large amounts of cyclodextrin may actually inhibit take-up of active agent.

In practice, a balance of these various considerations may apply in any particular

circumstance. The water-swellable hydrogel body may be formed of a cross-linked or uncross- linked hydrogel which maintains its integrity when swollen in water. The hydrogel is preferably a polyurethane formed by the reaction of a polyol and an isocyanate. Cross- linked hydrogels are typically formed by the reaction of a diol, a triol and a diisocyanate as described in patent specifications EPOO 16652, EPOO 16654 and WO/00/32171. Linear uncross-linked hydrogels are typically formed by the reaction of a diol (especially decane diol) and an isocyanate, especially as described in patent specification WO/2004/029125).

The present invention allows the delivery of hydrophobic pharmaceutically active agents, which have low water-solubility. Generally speaking, low water solubility means that the active agent is insufficiently soluble in water to allow a dose (e.g. or daily dose) for a patient to be loaded into the hydrogel body and delivered therefrom into an aqueous biological system. For example, a typical daily dose of progesterone is lOOmg, of 17 α- hydroxy progesterone caproate is 35mg and of fluconazole is lOOmg. Thus, in practice the water -solubility of an active agent is linked to the desired daily dosage to be administered. In terms of absolute solubilities, this generally means that a drug is insufficiently soluble in water if its solubility is less than lOmg, particularly 5mg, especially 2mg and often less than lmg per ml of water at 25⁰C. However, as indicated above, the acceptable solubility limits would generally be related to the dose to be administered.

Pharmaceutically active agents which are generally speaking too insoluble to be delivered from a hydrogel delivery vehicle (and therefore which require to be formulated according to the present invention) include one or more of the following:

antifungals such as amphotericin, clotrimazole, econazole nitrate, fluconazole, griseofulvin,

itraconazole, ketoconazole, miconazole nitrate; hormones such as progesterone, 17α

hydoxyprogesterone, dydrogesterone, medroxyprogesterone acetate, nandrolone, norgestrel, oestradiol, testosterone; anti-epileptics such as carbamazepine, clonazepam, phenobarbitone, niethylphenobarbitone; antiprotozoal agents such as metronidazole; antiviral agents such as aciclovir; anxiolytics and antipsychotics such as diazepam, droperidol, haloperidol, lorazepam, midazolam; cardiovascular agents such as glyceryl trinitrate, nifedipine; corticosteroids such as betamethasone, dexamethasone, hydrocortisone, prednisolone; prostaglandins such as dinoprostone, alprostadil, misoprostol.

An amount of cyclodextrin should be included in the hydrogel delivery device which is sufficient to give a solubility of active agent in water/cyclodextrin or water/ethanol/cyclodextrin solution which is generally in excess of lOmg, preferably 20mg, particularly 50mg and especially lOOmg per ml of water/cyclodextrin or water/ethanol/cyclodextrin solution (ethanol content up to 95%, particularly up to 75%, especially up to 50% of water-ethanol content by volume).

Usually, the molar ratio of active agent to cyclodextrin ranges from 1 to 0.2 (particularly 0.5) up to 1.5 (particularly 4.0) in the delivery device.

The hydrogel delivery device of the present invention will generally be sold in an essentially dry state (water content less than 15% and often less than 10% by weight). The cyclodextrin will usually comprise less than 50% by weight of the delivery device. Usually, the cyclodextrin content will be in excess of 0.01, particularly 0.1 and especially 1% by weight. The upper limit may be 10%, especially 30% and particularly 45% by weight.

Usually, the pharmaceutically active agent will be present in an amount of up to 30% (typically up to 20%) of weight of the delivery device.

Embodiments of the present invention will now be described by way of Example only. In the "normalised % release" profiles, the final amount released in the dissolution profile is taken as 100% for each release profile. This highlights differences in the shape of

the release profiles.

The following figures illustrate the invention: Figure 1 shows a β-cyclodextrin molecule with seven glucose units;

Figures 2 and 3 show dissolution profile and normalised % release of 15 mg of progesterone from cross-linked pessaries with and without cyclodextrin;

Figures 4 and 5 show dissolution profile and normalised % release of 17-AHPC in

55% ethanol at 37°C from cross-linked pessaries loaded with and without cyclodextrin;

Figures 6 and 7 show dissolution profile and normalised % release of 35 mg of 17-

AHPC in 6% sodium lauryl sulphate at 37⁰C from cross-linked pessaries loaded with and

without cyclodextrin;

Figures 8 and 9 show dissolution profile and normalised % release of 15 mg of

fluconazole in water at 37°C from cross-linked pessaries loaded with (FU04001) and without

cyclodextrin (FU04012). Loading solution was water;

Figures 10 and 11 show dissolution profile and normalised % release of fluconazole

in water at 37°C from cross-linked pessaries loaded with 15 mg (FU04012) and 100 mg

(FU04013) of the drug. Loading solution was water;

Figures 12 and 13 show dissolution profile and normalised % release of 50 mg of

fluconazole in water at 37°C from cross-linked pessaries loaded with various concentrations

of HPBCD (FU04015, FU04016 and FU04017) and without HPBCD (FU04014). Loading solution was 25% ethanol;

Figures 14 and 15 show dissolution profile and normalised % release of 50 mg of

fluconazole in water at 37°C from cross-linked pessaries loaded with various concentrations

of HPBCD (FU04019, FU04020 and FU04021) and without HPBCD (FU04018). Loading solution was 50% ethanol;

Figures 16 and 17 show dissolution profile and normalised % release of 50 mg of

fluconazole in water at 37°C from linear pessaries loaded with various concentrations of HPBCD (FU05002, FU05003 and FU05004) and without HPBCD (FU05001). Loading solution is 50% ethanol;

Figures 18 and 19 show dissolution profile and normalised % release of 50 mg of

fluconazole in water at 37⁰C from cross-linked pessaries loaded with 50 mg of HPBCD.

Loading solution was either water (FU04022), 25% ethanol/water (FU04016) or 50% ethanol/water (FU04020);

Figures 20 and 21 show dissolution profile and normalised % release of 100 mg of

fluconazole in water at 37°C from cross-linked pessaries loaded with various concentrations

of HPBCD (25 mg: FU04005, 100 mg: FU04006 and 250 mg: FU04007) over a period of 3 hours. Loading solution was 50% ethanol;

Figures 22 and 23 show dissolution profile and normalised % release of 50 mg of

fluconazole in water at 37°C from cross-linked pessaries loaded with various concentrations

of SBECD (FU04045, FU04046, FU04047) and without SBECD (FU04044). Loading solution was 50% ethanol;

1. ACTIVE AGENTS

1.1. Progesterone (PR)

The female sex hormone, progesterone, is a natural progestin that induces secretory changes in the lining of the uterus, which are essential for successful implantation of a fertilized egg. Progesterone has a molecular weight of 314.5. It is practically insoluble in water but soluble in alcohol.

1.2. 17α-Hydroxyprogesterone Caproate (17-AHPC)

17α-Hydroxyprogesterone is hormonally inactive. Its esterification with acetate leads to weak progestogenic activity, while esterification with caproate leads to a highly active progestin, which is clinically useful for intramuscular injection, and is found in the circulation as unmetabolised ester (Schindler et al. 2003).

17α-Hy droxyprogesterone caproate (17- AHPC) is a synthetic hormone similar to the natural female sex hormone progesterone. It is a white or creamy-white, crystalline powder. It is also odourless or having a slight odour. 17- AHPC has a molecular weight of 428.6. It is insoluble in water; however, it is soluble in ether and ethanol and slightly soluble in benzene. 17-AHPC should be protected from light (Reynolds, 1996; Sweetman, 2005).

17- AHPC is used in the prevention of spontaneous abortion in women with a history of recurrent miscarriage (habitual abortion). It is virtually inactive when given by mouth but works as a long-acting progestin when administered intramuscularly. Suggested doses have been 250 to 500 mg weekly by intramuscular injection during the first half of pregnancy.

1.3. Fluconazole (FU)

Fluconazole is a triazole antifungal agent, which is used for superficial mucosal (oropharygeal, oesophageal or vaginal) candidiasis as well as for fungal skin infections. It is also used for the treatment of systemic infections including systemic candidiasis, coccidioidomycosis and cryptococcosis (Sweetman, 2005).

Fluconazole is well absorbed following oral administration. In normal volunteers, the bioavailability after oral intake is 90% or more as compared with intravenous administration (Sweetman, 2005). Its plasma half-life is about 30 hrs (Gennaro, 2000). Fluconazole is a white crystalline solid. It has a molecular weight of 306.27. Its solubility in water is 8 mg/ml at 37C and in alcohol is 25 mg/ml at room temperature (Gennaro, 2000; Fluconazole, 2003).

2. METHODOLOGY

2.1. Solubility Studies

Solubility of various drug, water, ethanol and cyclodextrin combinations was determined by adding different drug quantities to prepared solvents and visually examining maximum solubility at room temperature. Soluble samples were incubated at 4°C, 25⁰C and 37°C overnight and their solubility was visually re-examined.

2.2. Manufacturing conditions

The linear polymer, batch LP04010, was manufactured according to patent WO2004029125 and sliced to produce pessaries of 0.8 mm thickness. The polymer is made of polyethylene glycol 8000/1, 10-decanediol/dicyclohexylmethane-4,4-diisocyanate (4.42/45.48/50.10 mol%).

Cross-linked hydrogel polymer batches, 0628:04103P, 0628:06503P and 0628:03304P of 0.8 mm thickness, were manufactured according to the general methodology set out in patent specifications GB2047093 and GB2047094. Briefly, the polymer is prepared by melting polyethylene glycol, adding hexanetriol in which ferric chloride is dissolved and then adding dicyclohexylisocyanate. The mixture is then cured at 95⁰C in a mould. The stoichiometric ratio of the components is calculated depending on the molecular weight of the

polyethyleneglycol. The batches were tested for equilibrium solvent uptake (swelling) after 24 hours in water at 25⁰C and 25% ethanol/water at 4⁰C. The batches were tested by accurately weighing 10 individually numbered pessaries then swelling in 300ml of demineralised water at 25 C or 300ml of 25% ethanol/water at 4⁰C as appropriate for 24hours. The pessaries were removed from the solution, blotted dry and reweighed and the swelling calculated as:

% Swelling = Weight of swollen polymer - Weight of dry polymer x 100

Weight of dry polymer

The results are detailed in Table 3.

Table 3. Testing results of batches 0628:04103P, 0628:06503P and 0628:03304P

460 unit lots of polymer 0628:04103P, 1000 units of polymer 0628:06503P and 1000 units of 0628:03304P were purified and drug loaded according to the methodology outlined below.

Purification:

Pessary slices were purified by washing in a rotating wash vessel (a Duran) on a Stuart Scientific Roller Mixer SRTl at the study temperature in a Gallenkamp Cooled Incubator and in a weight of solution equivalent to 10x weight of dry polymer using 3 cycles: firstly in purified water for 6-8 hrs, followed by a second purified water wash for 16-20 hrs and lastly a 25% ethanol/water wash for 6-8 hrs. The purified, blank polymers were dried using a Heraeus vacuum oven or Buchi R-111 Rotavapor at ambient temperature.

Purification of the linear polymers was not performed. Loading:

The hydrogels were drug loaded by swelling in a weight of drug loading solution equivalent to the total expected uptake of the polymer.

A number of dry hydrogel inserts (25 - 100 units) were immersed in a solution containing the active drug under slow rotation on the Stuart Roller Mixer. The loading solution varied between water, water/HPBCD, water/ethanol and water/ethanol/HPBCD. The loading temperature was also varied, depending on solubility studies.

The following equations were used to find the theoretical weight of the drug and the theoretical loading solution.

Theoretical weight of drug required = no. of units x target potency

Theoretical weight of the loading solution required = wt of units x (swelling factor/ 100)

Drying:

Swollen drug loaded inserts were dried either in the Heraeus vacuum oven or by using Buchi R-Hl Rotavapor at ambient temperature. Progesterone and 17- AHPC units were protected from light while drying.

Details of batch manufacturing for progesterone, fluconazole and 17α- hydroxyprogesterone caproate loaded hydrogel are presented below. 2.2.1. Progesterone

Two progesterone batches, with and without HPBCD, were loaded into hydrogel pessaries. Manufacturing details are given in Table 4 (% weight is weight per total weight of drug, HPBCD and polymer):

Table 4

2.2.2. 17α-Hydroxyprogesterone Caproate

Eight batches of 17α-hydroxyprogesterone caproate were loaded to asses the effect of HPBCD concentration and solvent concentration on loading and release from hydrogel pessaries. Manufacturing details are given in Table 5 (%weight is weight per total weight of drug, HPBCD and

polymer):

Table 5

ON

2.2.3. Fluconazole

Batches of fluconazole were manufactured to assess the effect of drug concentration, HPBCD concentration, temperature and loading solvent on the loading and release from hydrogel pessaries. Details of batch manufacturing are given in Table 6 (%weight is weight per total weight of drug, HPBCD and polymer):

Table 6

VO

Four batches were manufactured to assess the effect of HPBCD concentration on the loading and release of 50 mg of fluconazole from hydrogel linear pessaries. Details of batch manufacturing are given in Table 7 (% weight is weight per total weight of drug, HPBCD and polymer):

Table 7

K)

O

Four batches of fluconazole were manufactured to assess the effect of drug and SBECD concentrations on the loading and release from hydrogel cross-linked pessaries. Details of batch manufacturing are given in Table 8 (% weight is weight per total weight of drug, SBECD and polymer):

Table 8

* %LOD was not carried out because drying time was more than 3 days. Therefore, the pessaries were assumed to be dry.

2.3. Analytical Testing

Dried, loaded units were sampled and analysed for LOD (loss on drying) by thermogravimetric analyser - Perkin Elmer TGA 7 using to determine the endpoint of the drying process.

Six dried units from each batch were tested for dissolution in water or ethanol/water at 37°C, using an Icalis automated dissolution apparatus.

Because 17- AHPC is water-insoluble, dissolution was carried out in 55% ethanol:water and in 6% sodium lauryl sulphate.

3. RESULTS

Table 9 shows the swelling factors of the polymers in different swelling media at different temperatures.

Table 9. Swelling factors of A0628:04103P, 0628:06503P, 0628:03304P and LP04010.

3.1. Progesterone

3.1.1. Progesterone Solubility

Table 10 shows the solubility of progesterone (PR) in 1% β-cyclodextrin, with and

without ethanol.

Table 10. Solubility results of progesterone (PR) in 1% β-cyclodextrin, with and without

ethanol.

Table 11 shows the solubility of PR in various concentrations of HPBCD in water, while Table 12 shows the solubility of PR in HPBCD/PEG 400/water. Table 11. Solubility of progesterone in HPBCD/water.

Table 12. Solubility of progesterone in HPBCD/PEG 400/water.

3.1.2. Progesterone Dissolution

Fig. 2 presents the dissolution of progesterone-HPBCD (batch PR04001) and progesterone alone (batch PR04003) from 0.8 x 10 x 30 mm pessaries over a period of 4 hrs. Figure 3 shows normalised % release of 15 mg of progesterone from cross-linked pessaries with and without cyclodextrin.

Table 13. Target HPBCD and progesterone content and actual progesterone content after 7-

hr dissolution in water at 37°C.

3.2. 17α-Hydroxyprogesterone Caproate

3.2.1. 17α-Hydroxy progesterone Caproate Solubility

The solubility of 17α-hydroxyprogesterone caproate in water is shown in Table 14, in ethanol is shown in Table 15, in 50% ethanol: water is shown in Table 16 and in ethanol/water/HPBCD is shown in Table 17.

Table 14. Solubility of 17-α-hydroxyprogesterone caproate in 1 g of water.

Table 15. Solubility of 17-α-hydroxyprogesterone caproate in 1 g of ethanol.

Table 16. Solubility of 17-α-hydroxyprogesterone caproate in 50% ethanol: water.

Table 17. 17-α-Hydroxyprogesterone caproate - solubility in ethanol/water/HPBCD

3.2.2. 17α-Hydroxyprogesterone Caproate Dissolution

Figures 4 and 5 show the dissolution of 17-AHPC (35 mg) in 55% ethanol / 45%

water (v/v) at 37°C, with and without various concentrations of HPBCD over a period of 4

hrs. Loading solution was either 70% ethanol (batches PC04001, PC04002, PC04003 and PC04004) or 50% ethanol (batches PC04005, PC04006, PC04007 and PC04008). Table 18. Target HPBCD and 17-AHPC content and actual 17- AHPC content after 4-hr

dissolution in 55% ethanol at 37°C.

Figures 6 and 7 show release of 17-AHPC from the pessaries using 6% SLS as a dissolution medium. Six batches were studied, namely PC04001, PC04002, PC04003, PC04005, PC04006 and PC04007. Dissolution was carried out for 17 days.

Table 19. Target HPBCD and 17-AHPC content and actual 17-AHPC content after 17-day

dissolution in 6% sodium lauryl sulphate at 37°C.

3.3. Fluconazole

3.3.1. Fluconazole Solubility

Preliminary solubility work was carried out looking at various solvent compositions to determine suitable conditions for loading fluconazole in polymer units. Table 20 shows FU solubility in various solutions containing HPBCD, ethanol and water. Table 21 shows FU solubility in solutions of ethanol/water and SBECD/ethanol/water.

Table 20. Solubility of fluconazole with HPBCD in different solvents.

K>

Table 21. Solubility of FU in SBECD/ethanol/water

O

3.3.2. Fluconazole Dissolution

Figures 8 and 9 show the dissolution of fluconazole (15 mg) in water at 37°C,

over a period of 4 hrs. Pessaries were loaded with and without HPBCD. Loading solution was water and polymers were cross-linked.

Table 22. Target HPBCD and fluconazole content and actual fluconazole content

after 4 hour dissolution in water at 37°C.

Figures 10 and 11 studies the release of different concentrations of fluconazole from cross-linked pessaries. No cyclodextrin was loaded in the polymers.

Table 23. Target HPBCD and fluconazole content and actual fluconazole content

after 3 hour dissolution in water at 37°C.

Figures 12 and 13 shows the dissolution of fluconazole (50 mg) in water at

37⁰C, over a period of 3 hrs. Pessaries were loaded with and without various

concentrations of HPBCD. Loading solution was 25% ethanol and polymers were cross-linked. Table 24. Target HPBCD and fluconazole content and actual fluconazole content

after 3 hour dissolution in water at 37°C.

Figures 14 and 15 show the dissolution of fluconazole (50 mg) in water at

37°C, over a period of 2.5 hrs. Pessaries were loaded with and without various

concentrations of HPBCD. Loading solution was 50% ethanol and polymers were cross-linked.

Table 25. Target HPBCD and fluconazole content and actual fluconazole content

after 2.5 hour dissolution in water at 37°C.

Figures 16 and 17 show the dissolution of fluconazole (50 mg) in water at

37⁰C, over a period of 5 hrs. Pessaries were loaded with and without various

concentrations of HPBCD. Loading solution was 50% ethanol and polymers were linear. Table 26. Target HPBCD and fluconazole content and actual fluconazole content

after 5 hour dissolution in water at 37⁰C.

Figures 18 and 19 compare the effect of loading solution on the entrapment of 50 mg of fluconazole into cross-linked polymers. Dissolution was carried out in

water at 37⁰C over a period of 3.5 hours. Fifty mg of HPBCD were also loaded with

each pessary. Loading solution was water (FU04022), 25% ethanol (FU04016) or 50% ethanol (FU04020). Polymers were cross-linked.

Table 27. Target HPBCD and fluconazole content and actual fluconazole content

after 3.5 hour dissolution in water at 37°C.

Figures 20 and 21 show the dissolution of 100 mg of fluconazole loaded with 25 mg (FU04005), 100 mg (FU04006) and 250 mg (FU04007) of HPBCD into hydrogel cross-linked polymers. Loading solution was 50% ethanol and polymers

were cross-linked.

Table 28. Target HPBCD and fluconazole content and actual fluconazole content

after 3 hour dissolution in water at 37°C.

Figures 22 and 23 show the dissolution of 50 mg of fluconazole loaded with 0 mg (FU04044), 12.5 mg (FU04045), 50 mg (FU04046) and 125 mg (FU04047) of SBECD into hydrogel polymers.

Table 29. Target SBECD and fluconazole content and actual fluconazole content

after 4 hour dissolution in water at 37⁰C.

4. DISCUSSION

4.1. Progesterone

4.1.1. Solubility of Progesterone

Progesterone (PR) (0.1 mg/ml) was insoluble when mixed with 1 ml of 1% β-

cyclodextrin (Table 10). Ethanol was added to improve solubility, although β- cyclodextrin is practically insoluble in ethanol, PR is soluble (1 in 8). However,

0.078 mg/ml of PR was still insoluble in 1 ml of a 25% ethanol : 75% of a 1 % β-

cyclodextrin solution, even after 5 min sonication at room temperature.

Dissolving PR in a solution of HPBCD/water showed that the solubility is dependent on the molar ratio of PR to HPBCD. As the concentration of HPBCD increases, solubility of PR improves (Table 11).

At low molar ratios of HPBCD (1:1.13 and 1 :2.3, PR:HPBCD), PR was insoluble even with 60 min sonication at room temperature.

Solubility of PR was much improved when PR:HPBCD molar ratio approached 1:3.4. Sonication at room temperature for 60 min also helped in providing clear solutions.

Solubility of PR in HPBCD/PEG 400/water was not improved, even at high HPBCD concentrations and with 60 min sonication at room temperature (Table 12).

The results of Table 11 show that 5 mg of progesterone was soluble in 1 ml of 75, 100 and 125 mg/ml concentration of HPBCD and did not precipitate or crystallise

after overnight storage at 4°C. A concentration of 75 mg/ml of HPBCD was chosen

as the loading solution.

4.1.2. Loading of Progesterone

Loading of progesterone into polymers at 4⁰C resulted in complete absorption of the loading solution by the pessaies. Without HPBCD, 4 mg of progesterone was insoluble in water and resulted in particles on the walls of the duran, whereas PR/HPBCD was soluble (see Table 11). hi both cases, 60 min sonication at room temperature was applied. The loaded, dried units were approximately 10% larger in size, even though LOD was 0.5%, implying that progesterone HPBCD were loaded into the polymers.

Fig. 2 shows the release profile of the hydrophobic drug, progesterone, with and without HPBCD (batches PR04001 and PR04003, respectively), from cross- linked pessaries in water at 37⁰C over 4 hours. Progesterone targeted dose in both batches was 4 mg/unit.

Fig. 2 clearly shows the effect of HPBCD on the release of the hydrophobic drug, progesterone, in water, where almost the entire loaded drug in batch PR04001 was released while only a quarter of the drug was released from pessaries without HPBCD.

In addition of being a solubilising agent, HPBCD played other roles, such as loading and releasing of progesterone from the pessaries.

4.2 17α-Hvdroxyprogesterone Caproate

4.2.1. Solubility of 17α-Hydroxy progesterone Caproate (17AHPC)

Solubility of 17 AHPC was carried out in water (Table 14), ethanol (Table 15), 50% ethanol / water (Table 16) and in ethanol/water/HPBCD (Table 17). 17 AHPC is insoluble in water but soluble in ethanol at concentrations of more than 200mg/ml with sonication.

Thirty five mg of 17- AHPC was also insoluble in 50% ethanol: water (Table 16). Sixty min sonication at room temperature did not aid in dissolving the hydrophobic drug, 17-AHPC. HPBCD:ethanol: water was tested as the dissolving solution. Table 17 shows that 67 mg of 17-AHPC is soluble in 1 g of 70% ethanol: water, containing 67 mg of

HPBCD. However, the sample was crystallised when incubated at 4°C and at 25°C

for 16 hrs and 30 min. Overnight incubation at 37°C resulted is a soluble solution.

Therefore, HPBCD with 70% ethanol was used as the loading medium to load

17-AHPC in hydrogel pessaries at 37⁰C.

4.2.2. Loading of 17α-Hydroxyprogesterone Caproate

All loaded pessaries after drying were covered with white powder, which is most probably 17-AHPC. This infers that the drug was not completely loaded in the pessaries.

Moreover, %LOD of the pessaries loaded with 70% ethanol ranged from 2.2 to 3.8%, and those loaded with 50% ethanol ranged from 1.5 to 6.5, thus the polymers were not dry, even though the drying time was 64 hrs and 23 hrs 20 mins respectively (Table 5).

Fig. 4 shows the dissolution of 17-AHPC in a medium of 55% ethanol / 45%

water (v/v) at 37°C over a period of 4 hrs.

Loading 17-AHPC with HPBCD in 70% ethanol solution showed a tremendous effect on loading and release of the drug. A large amount of 17-AHPC of batch PC04001 (35 mg of 17-AHPC without HPBCD) was not truly loaded in the polymer since 62% (24 mg) of the drug was released within the first 10 minutes. Release slowed down after 14 minutes and complete dissolution is seen after 90 mins. On contrary, PC04002, PC04003 and PC04004, which were loaded with 50, 100 and 150 mg of HPBCD, respectively, showed a more controlled release for 2 hrs. Compared with 62% release at 10 min in PC04001, the burst release was less with HPBCD-loaded polymers: 24% or 8 mg (PC04002), 22% or 6 mg (PC04003) and 18% or 5 mg (PC04004). As the amount of loaded HPBCD increased, total drug released at 4 hrs decreased: 34 mg (PC04002), 27 mg (PC04003) and 29 mg (PC04004) of 17- AHPC. It might mean that either the pessaries did not release the entire drug or 17- AHPC was not completely loaded in the pessaries, as HPBCD occupied most of the space within the pessaries.

Two-hour controlled release is observed with pessaries loaded with 17- AHPC and HPBCD, using 70% ethanol.

Loading using 50% ethanol was assessed, even though the drug is not soluble in the loading solution. Fig. 4 shows the following results:

• At 4 hours, the pessaries did not release the targeted dose of 17-AHPC, i.e. 35 mg. PC04005 (no HPBCD) released 22 mg, PC04006 (50 mg of HPBCD) released 17 mg, PC04007 (100 mg of HPBCD) released 14 mg, and PC04008 (150 mg of HPBCD) released 11 mg. It is clear that as the amount of HPBCD increases, which competes for drug entrapment, the total amount of drug released at 4 hrs decreases. It is possible that the drug was not completely loaded in the polymer, especially it is not soluble in 50% ethanol.

• HPBCD-free polymers showed higher burst release as compared with HPBCD-loaded pessaries. At 10 mins, 13 mg or 59% was released from PC04005, while 8 mg (47%), 6 mg (44%) and 4 mg (37%) was released from PC04006, PC04007 and PC04008, respectively.

• Minimal release is noted after 95 min dissolution. Pessaries loaded with 70% ethanol showed more loaded drug and better release profile than those loaded with 50% ethanol.

An alternate method using 6% sodium lauryl sulphare (SLS) was used to study the

dissolution of 17-AHPC from pessaries. Six batches were studied at 37°C: PC04001,

PC04002, PC04003, PC04005, PC04006 and PC04007. The dissolution analysis was carried on 1 or 2 units only.

Pessaries loaded with 17-AHPC using 70% ethanol entrapped and released more drug than pessaries loaded using 50% ethanol (Fig. 6).

HPBCD-free polymers (PC04001 and PC04005) released drugs faster than HPBCD-loaded pessaries (PC04002, PC04003, PC04006 and PC04007). At 3 days, 29 mg of 17-AHPC was released from PC04001, while PC04002 and PC04003 released 13 and 10 mg, respectively (Fig. 6). Similarly, PC04005 released 22 mg, where PC04006 and PC04007 released 12 and 9 mg, respectively.

None of HPBCD-loaded pessaries showed 35 mg release of 17-AHPC within one week. At 7 days, the best release profile was obtained with PC04002, where 20 mg was released. Batch PC04001 (HPBCD-free pessaries) showed 34 mg release of 17-AHPC; however, most of the drug was burst released.

A slow release profile is observed over the 17-day period. At the end of the dissolution, PC04002 released 30 mg, while PC04003 released 25 mg. Batches PC04006 and PC04007 released even lower amounts of 17-AHPC at 17 days: 19 and

15 mg, respectively.

The higher amount of drug dissolution from HPBCD-free pessaries (batches PC04001 and PC04005) could be due to unloaded 17-AHPC accumulated at the surface of the polymer.

4.3. Fluconazole

4.3.1. Solubility of Fluconazole

Solubility of fluconazole in a medium of water/HPBCD at room temperature is summarised in Table 20. The results show that up to 25 mg of fluconazole is soluble in 1 ml of water, containing 125 mg of HPBCD. No precipitation or

crystallisation of drug was formed after overnight incubation at 37°C.

Table 20 also shows the solubility of 100 mg of fluconazole in 50% ethanol/water (w/w), containing 0 to 200 mg of HPBCD. Sonication helped solubilising 100 mg of fluconazole in 2 g of 50% ethanol, with and without HPBCD.

Incubating overnight at 25⁰C and 37⁰C resulted in clear, soluble solutions.

A higher weight of the drug (200 mg) was also tested in a solution of 50%

ethanol/water (w/w), containing 200 mg of HPBCD. The sample was soluble at 25°C

and therefore, batches FU04005, FU04006 and FU04007 of 100 mg of fluconazole were loaded in 50% ethanol/water at 25°C.

SBECD was used as an alternative to HPBCD. Solubility of 100 mg of fluconazole was tested with different concentrations of SBECD (0 to 200 mg) using 50% ethanol/water (w/w). Table 21 shows that 100 mg of fluconazole was soluble with SBECD (10 to 200 mg) using 2 g of 50% ethanol; the sample stayed clear after

incubation overnight at 4°C, 25°C and 37°C. Without SBECD, the sample required

sonication to improve solubility.

4.3.2. Loading of Fluconazole with HBPCD in Cross-Linked Polymers

The different batches of fluconazole-loaded pessaries, together with HPBCD and SBECD potencies, the loading media and temperatures are listed in Tables 6, 7 and 8.

Because the solubility of FU was improved using HPBCD, ethanol and water, different concentrations of fluconazole (10 to 100 mg) were loaded into cross-linked polymers, using different loading media. 50mg of fluconazole was also loaded into linear polymers using HPBCD.

The effect of SBECD on the loading and release of fluconazole was also

studied.

15 mg of Fluconazole

Twenty four hour dissolution was carried out on hydrogel polymer units loaded with 15 mg of fluconazole per pessary with (FU04001) and without HPBCD

(FU04012) in water at 37°C. The dissolution results reported in Fig. 8 are for 240 minutes, where maximum release of fluconazole was observed, although the dissolution was carried out for 24 hours.

Fig. 8 clearly shows a controlled, steady release of fluconazole from the pessaries over 1 hour 30 min. The results indicate that the hydrophobic drug, fluconazole, was loaded into the hydrogel pessaries. When fluconazole was loaded into polymers without HPBCD, similar results were obtained.

15 mg vs. 100 mg of Fluconazole

Fig. 10 shows the ability of fluconazole in loading into cross-linked hydrogel polymers. Fifteen mg of fluconazole was easily loaded. However, when target potency of 100 mg was loaded, only one-third of the target potency was actually entrapped in the cross-linked polymers. Moreover, a burst release of 15 mg was noticed within the first 4 min of FU04013, which targeted 100 mg of fluconazole.

50 mg of Fluconazole

With better solubility of fluconazole, using ethanol/HPBCD/water, concentrations of 50 mg were loaded into cross-linked polymers at 25 °C (loading solution was 25% ethanol: water). Fig. 12 shows a quick release during the first few minutes, followed by a controlled release of fluconazole from the polymers, with and without HPBCD. The amount of drug released was quite similar in all the batches, implying that almost equal amount of fluconazole was loaded in all batches. A trend existed of decreased initial drug release with increasing concentration of HPBCD. The batch with no HPBCD (FU04014) had 52% of drug released in the first 6 minutes, whereas the batch with 125 mg of HPBCD (FU04017) released only 34% in the first 6 minutes. This indicates that HPBCD improves penetration of fluconazole into the hydrogel polymer, leaving less surface coating.

However, actual flucoanzole content in all batches that used 25% ethanol as the loading solution was around 40 mg (Table 24), i.e. 20% less than the target drug dose.

Release of 50 mg of fluconazole from cross-linked polymers using 50% ethanol as the loading solution is shown in Fig. 14. Similar results were obtained for the different batches, with and without HPBCD, except for batch FU04021, which contained the highest amount of HPBCD (125 mg). FU04021 showed better controlled release and less amount of fluconazole content at the end of the dissolution period, i.e. 2.5 hrs (Table 25).

Controlled, slow release is shown for the first 90 min, where most of the drug was released (FU04018: 48.732 mg, FU04019: 50.719 mg, FU04020 47.775 mg and FU04021 : 40.750 mg), followed by a plateau.

50 mg of Fluconazole: Effect of Loading Solution

The effect of loading solution on the entrapment of 50 mg of fluconazole, using 50 mg of HPBCD, was studied. Fig. 18 clearly shows that as the amount of ethanol in the loading solution increases, loaded fluconazole also increases. This is due to the increased solubility of the drug in the loading solution. At 3.5 hrs of dissolution, 20.58 mg of fluconazole was released from FU04022 (water is the loading solution) (Table 27). The released amount doubled to 39.67 mg when water was replaced by 25% ethanol (FU04016). Using 50% ethanol (FU04020), the target content of fluconazole of 50 mg was loaded into the pessaries and released within 3.5 hrs.

Moreover, release was better controlled when the loading solution was 50% ethanol (FU04020), where 24% of the actual fluconazole content was released at 8 min, as compared to 50% and 48% of drug released from pessaries having water and 25% ethanol as the loading solutions, respectively (Fig. 19).

Results from Fig. 18 infer that drug solubility in the loading solution is a major factor in controlling the loading of drug into cross-linked polymers. Unsolublised drug particles would not be able to cross the polymeric layers of the pessary but they retain at the surface of the polymer. Thus a large amount of drug would not load, and what loaded is quickly released.

In general, batches loaded with 50 mg of fluconazole were better in appearance than the 15 mg batches, which were loaded in water, with and without HPBCD (FU04001 and FU04012, respectively). Powder covering the pessaries was either minimal or not present.

100 mg of Fluconazole

Fig. 20 presents the release profiles of different fluocnazole batches of lOOmg dose, with different concentrations of HPBCD. As the amount of HPBCD in the formulation increases, the amount of released fluconazole decreases (Table 28). It is worth noting that the amount of loaded drug and HPBCD in batch FU04006 is nearing the weight of a pessary (200 mg vs. 241 mg), while the weight of drug and HPBCD in FU04007 is much larger than a pessary weight (350 mg vs. 241 mg) (Table 6). Therefore, the large amount of HPBCD and fluconazole would not find a room within the polymeric layers to load. This might explain the low amount of released fluconazole from batch FU04007.

Fig. 20 also shows instantaneous release in the first few minutes of dissolution. As the amount of HPBCD increases, burst release decreases. At 4 min, 25 mg was released from FU04005, 19 mg from FU04006 and 11 mg from FU04007.

Moreover, Fig. 20 shows that the highest concentration of HPBCD entrapped in batch FU04007 (250 mg) presented better controlled release than batch FU04005, which contained the lowest amount of HPBCD (25 mg). The reason could be that HPBCD enhances the solubility of fluconazole in loading solution due to which fluconazole can be loaded deep into the matrix, which ultimately controls the release rate of fluconazole. Fluconazole in the insoluble form will be retained either on or near the surface and will be released within few minutes of run.

Thus, HPBCD not only controls the loading of flucoanzole into polymers, but also governs its release rate.

In contrast to the 50 mg batches, all fluconazole batches of 100 mg were powdery in appearance and a large proportion of powder was present inside the packaging container. 4.3.3. Loading of Fluconazole with HBPCD in Linear Polymers

Loading 50 mg of fluconazole and different concentrations of HPBCD into linear polymers showed similar release patterns among the batches studied (Fig. 16). Within the first 4 min, 7 to 8 mg of drug was released from the different batches. Within 90 min, 46 mg of fluconazole was released from HPBCD-free polymer (FU05001) and 44 mg was released from HPBCD-loaded polymers; thus the difference among the polymers is small (see also Table 26). Nonetheless, the presence of the HBPCD alters the release profile (see Fig.17), increasing amounts of HBPCD tending to slow the release.

More differences are seen between HPBCD-free polymers and HPBCD- loaded polymers during 6 and 60 min. For example, at 14 min, 23 mg of fluconazole was released from FU05001 while 16 mg was released from FU05004.

The similar dissolution results, with and without HPBCD, is due to the type of the linear polymer used. These linear polymers are capable to load hydrophobic drugs, such as fluconazle, without the need of HPBCD, which could be due to the hydrophobic monomer, 1,10-decanediol, in the linear polymer.

It is worth mentioning that batch FU05004, which contained the largest amount of HPBCD (125 mg per pessary), showed the best appearance: clear and transparent pessaries. From another point of view, none of the linear polymers was covered with powder. Comparing release of 50 mg of fluconazole from the two types of polymers, cross-linked and linear (Figs. 14 and 16), release from cross-linked polymers is slower and better controlled than from linear polymers. For example, at 10 min:

Release from HPBCD-free pessaries, batch FU04018, is 11 mg and from FU05001 is l8 mg.

• Release from FU04019 is 11 mg and from FU05002 is 18 mg.

• As the concentration of HPBCD increases in both polymers, release differences decrease. FU04020 releases 14 mg and FU05003 releases 14 mg.

• FU04021 releases 10 mg and FU05004 releases 13 mg.

In addition, as dissolution time increases, release differences between the two polymers decrease. At 40 min:

• Release from FU04018 is 35 mg and from FU05001 is 38 mg.

• Release from FU04019 is 36 mg and from FU05002 is 38 mg.

• Release from FU04020 is 36 mg and from FU05003 is 34 mg.

• Release from FU04021 is 29 mg and from FU05004 is 33 mg.

The slower release of fluconazole from cross-linked polymers could be related to the physical structure of the polymer. As their name suggest, layers of cross-linked polymers restrict drug movement and thus slow down and control release as compared to layers of linear polymers.

4.3.4. Loading of Fluconazole with SBECD in Cross-Linked Polymers

50mg of fluconazole were loaded in hydrogel pesaries, with and without

SBECD, in a loading medium of 50% ethanol/water at 25⁰C. Loaded pessaries were

patchy. Fig. 22 shows the dissolution of fluconazole from cross-linked polymers. The release results do not change as potency of SBECD changes. A 70-min controlled dissolution was observed in all batches, where 90% of the drug was released. Fifty percent release occurred between 24 and 27 min and complete dissolution was at 150 min.

Therefore, using 50% ethanol as the loading solution, the effect of SBECD is minimal.

Comparing the effect of HPBCD and SBECD on the release of fluconazole from cross-linked polymers (Figs. 14 and 22), minimal difference is shown in batches loaded with 12.5 mg and 50 mg of cyclodextrin. However, batches loaded with 125 mg of HPBCD and SBECD (batches FU04021 and FU04047, respectively) showed that HPBCD better controls drug release from cross-linked polymers. For example,

• at 20 min, release from FU04021 is 17 mg and from FU04047 is 22 mg.

• at 40 min, release from FU04021 is 29 mg and from FU04047 is 38 mg.

• at 60 min, release from FU04021 is 36 mg and from FU04047 is 45 mg.

It might be explained that fluconazole fits in HPBCD hydrophobic cavity better than SBECD, and thus slowly released.

Claims

1. A pharmaceutical delivery vehicle, which comprises: a water-swellable hydrogel body containing a pharmaceutically active agent together with a pharmaceutically acceptable cyclodextrin.

2. A delivery vehicle according to claim 1 wherein the water-swellable hydrogel is a cross-linked hydrogel.

3. A delivery vehicle according to claim 2 wherein the water-swellable hydrogel is a cross-linked polyurethane.

4. A delivery vehicle according to claim 3 wherein the cross-linked polyurethane is the reaction product of a diol, a triol and an isocyanate.

5. A delivery vehicle according to claim 1 wherein the water-swellable hydrogel is an uncrosslinked hydrogel.

6. A delivery vehicle according to any preceding claim wherein the solubility of the pharmaceutically active agent is less than lOmg per ml. of water at 25°C.

7. A delivery vehicle according to claim 6 wherein the solubility of the pharmaceutically active agent is less than 5mg per ml. of water at 25°C.

8. A delivery vehicle according to any preceding claim wherein the pharmaceutically active agent is progesterone.

9. A delivery vehicle according to any of claims 1 to 7 wherein the pharmaceutically active agent is 17α-hydroxyprogesterone caproate.

10. A delivery vehicle according to any of claims 1 to 7 wherein the pharmaceutically active agent is fluconazole.

11. A delivery vehicle according to any preceding claim wherein the cyclodextrin is a substituted or unsubstituted β-cyclodextrin.

12. A delivery vehicle according to claim 11 wherein the cyclodextrin is hydroxypropyl-β-cyclodextrin.

13. A delivery vehicle according to any of claims 1 to 10 wherein the cyclodextrin has a solubility in the range one part cyclodextrin in 1 to 10 parts water at 25°C.

14. A pharmaceutical delivery vehicle according to any preceding claim wherein the cyclodextrin is present in an amount which provides a solubility of active agent in water/cyclodextrin solution which is in excess of lOmg per ml.

15. A pharmaceutical delivery vehicle according to any of claims 1 to 13 wherein the cyclodextrin is present in an amount which provides a solubility of active agent in water/ethanol/cyclodextrin solution which is in excess of lOmg per ml.

16. A pharmaceutical delivery vehicle according to claim 15 wherein the ethanol is up to 75% by volume of the water/ethanol content.

17. A pharmaceutical delivery vehicle according to any of claims 1 to 14 wherein the cyclodextrin comprises less than 50% by weight of the delivery device.

18. A pharmaceutical delivery vehicle according to any of claims 1 to 14 wherein the cyclodextrin comprises in excess of 1% by weight of the delivery device.

19. A pharmaceutical delivery vehicle according to any preceding claim having a water content less than 10% by weight.