Biodegradable compositions suitable for controlled release

Abstract

A special class of drug-depot forming triblock copolymers which are very suitable for the loading, containment and releasing of sensitive drugs such as proteins from biodegradable, injectable drug depots. How to visualize these depots for various imaging related purposes is described. A composition comprising a tri-block copolymer according to formula 1 B-A-B (1), wherein A stands for a linear poly-(ethylene glycol) block and wherein B stands for wherein B stands for a poly(lactide-co--caprolactone) block, wherein the hydroxyl end-groups of the tri-block copolymer are at least partially acylated with an optionally substituted acyl having 2 to 12 C-atoms, C-atoms of the substituents included; an active ingredient, preferably a pharmaceutically active ingredient and a solvent, wherein the block ratio of the tri-block copolymer, which ratio is defined as the ratio between the sum of the average molecular weight of the B-blocks and the sum of the average molecular weight of the A-block ranges from 1.4 to 3.5. This composition is suitable for controlled release of a pharmaceutically active ingredient.

Claims

1. A composition, comprising: a tri-block copolymer according to formula 1
B-A-B(1) wherein A stands for a linear poly-(ethylene glycol) block and wherein B stands for a poly(lactide-co--caprolactone) block, wherein a weight ratio of -caprolactone to lactide is from 1/1 to 9/1; wherein hydroxyl end-groups of the tri-block copolymer are at least partially acylated, wherein the degree of acylation is at least 75% and is at most 100%, wherein the acyl group is an acetyl group, propionyl group or a butionyl group, an active ingredient, and a solvent, wherein a block ratio of the tri-block copolymer, which ratio is defined as the ratio between a sum of a number average molecular weight of the B-blocks and a number average molecular weight of the A-block ranges from 1.6 to 2.6, wherein a number average molecular weight (Mn) of the linear poly-(ethylene glycol) block is at least 1250 Da and at most 2000 Da, and wherein concentration of the tri-block copolymer in the composition is at least 15% w/w to at most 50% w/w based on the amount of the solvent and active ingredient present in the composition.

2. A composition, comprising: a tri-block copolymer according to formula (1)
B-A-B(1) wherein A stands for a linear poly-(ethylene glycol) block and wherein B stands for a poly(lactide-co--caprolactone) block, wherein at least part of hydroxyl end-groups of the tri-block copolymer are covalently bound to a compound containing radiopaque atoms, wherein a weight ratio of -caprolactone to lactide is from 1/1 to 9/1, wherein a block ratio of the tri-block copolymer, which ratio is defined as the ratio between a sum of a number average molecular weight of the B-blocks and a number average molecular weight of the A-block ranges from 1.6 to 2.6, wherein a number average molecular weight (Mn) of the linear poly-(ethylene glycol) block is at least 1250 Da and at most 2000 Da, and wherein hydroxyl end-groups of the tri-block copolymer are at least partially acylated with the compound containing radiopaque atoms, wherein the degree of acylation is at least 75% and is at most 100%, wherein concentration of the tri-block copolymer in the composition is at least 15% w/w to at most 50% w/w based on the amount of the solvent and active ingredient present in the composition.

3. The composition according to claim 2, wherein the composition further comprises an active ingredient and a solvent.

4. The composition according to claim 1, wherein a concentration of the tri-block copolymer in the composition of the invention ranges from 21 to 35% w/w based on the amount of the solvent and the active ingredient present in the composition.

5. The composition according to claim 1, wherein the acyl group is an acetyl group or a propionyl group.

6. The composition according to claim 1, wherein the active ingredient is a pharmaceutically active ingredient chosen from the group of steroids, and non-steroidal anti-inflammatory drugs.

7. The composition according to claim 1, wherein the active ingredient are microparticles, nano-particles, microspheres containing drugs or imaging agents, or liposomes containing siRNA, miRNA, drugs or imaging agents.

8. The composition according to claim 1, wherein the active ingredient has a solubility of at least 20 g/ml in water measured at 20 C. and at 1 bar pressure.

9. The composition according to claim 1, wherein the solvent is water or an aqueous buffer solution.

10. The composition according to claim 1, wherein the composition has a gel temperature in the range from 25 to 35 C. and a phase separation temperature of at least 42.

11. The composition according to claim 1, wherein the tri-block copolymer has a gel window between 25 C. and 42 C.

12. The composition according to claim 1, wherein the acyl group is substituted with a radiopaque atom.

13. The composition according to claim 2, wherein the radiopaque atom is iodine.

14. The composition according to claim 1, wherein the composition is a pharmaceutical composition further comprising a pharmaceutically acceptable adjuvant, excipient or carrier.

15. The composition according to claim 1 for use as a medicament.

16. A process for the preparation of a composition according to claim 1 comprising the steps of synthesizing the tri-block copolymer; and mixing the tri-block copolymer with the active ingredient and the solvent.

Description

EXAMPLES

Measurement Methods

(1) The tri-block copolymer composition (PLCA/PEG ratio, cap/lac ratio and DM) was determined with proton nuclear magnetic resonance (.sup.1H NMR; Varian, 400 MHz), using deuterated chloroform as solvent and reference. From the integration of various proton signals (due to PEG and incorporated monomers), absolute number average molecular weights M.sub.n were obtained. Mn of the tri-block copolymer is the sum of the molecular weights of the central PEG block and the two polyester blocks (ratio of blocks determined with NMR).

(2) The block ratio as used herein is the weight ratio of the PLCA-blocks to the PEG-block (of known molecular weight) and can be calculated from grams of monomers (lactide+caprolactone) divided by grams of PEG used to synthesize the tri-block copolymer. Final tri-block composition after polymer purification was checked with .sup.1H NMR by comparing integrals of peaks due to PEG and incorporated ring-opened monomers. The weight ratio of -caprolactone to L-lactide can be calculated from .sup.1H NMR by comparing integrals of peaks due to ring-opened lactide and caprolactone. The integrals of peaks due to acyl endgroup and PEG block were used to calculate the degree of modification (range 0-2, which corresponds to a degree of acylation of 0 to 100%).

Example 1: Preparation of the Tri-Block Copolymer #33: Acylated (C2-Modified) PLCA-PEG-PLCA

(3) Polyethyleneglycol (PEG 1500, 12.5 g, 8.3 mmol) and ca. 100 ml toluene were charged into a 250 ml three-neck round-bottom flask equipped with a magnetic stirring bar. Using a Dean-Stark device with a cooler on top, 40 ml of toluene was distilled off to remove water (from PEG) azeotropically by heating at 150 C. at atmospheric pressure (1 bar) under nitrogen.

(4) After cooling down the solution to ca. 80-100 C., L-lactide (2.75 g, 19 mmol) and caprolactone (24.75 g, 217 mmol) were added. 40 ml of toluene was distilled off to dry the monomers by heating at 150 C. at atmospheric pressure. Ca. 20 ml of dry toluene was left in the flask for the polymerization.

(5) After cooling down the mixture to ca. 80-100 C., tin(II) 2-ethylhexanoate (0.25 ml) was added through one of the necks.

(6) Polymerization was carried out at 120 C. for 1 day under nitrogen atmosphere.

(7) After cooling down to room temperature, ca. 70 ml dichloromethane and 5.6 ml triethylamine (40 mmol) were added. Subsequently, 2.5 ml acetyl chloride (33 mmol) was added slowly to the stirred solution, which was cooled with an ice bath. The acylation reaction was continued for a few hours after which dichloromethane was removed by rotavap, and ethyl acetate (ca. 100 ml) was added to the residue. Triethylamine salt was removed by (paper) filtration and the polymer dissolved in the clear filtrate was precipitated by addition of a mixture of hexane and diethyl ether. At ca. 20 C. (in freezer) the polymer product separated as a waxy solid from which non-solvents could be decanted easily. Finally, the precipitated polymer was dried in vacuo. Yield: ca. 37 g. The PLCA/PEG block ratio was around 2.2; the -caprolactone/lactide ratio was 9/1.

Example 2. Preparation of the Tri-Block Copolymer #27

(8) In a manner similar to example 1 an acylated PLCA-PEG-PLCA tri-block was synthesized having a PLCA/PEG ratio of 1.8 and a caprolactone/lactide weight ratio 9/1. The differences with example 1 were: different PLCA/PEG ratio and different caprolactone/lactide weight ratio.

Example 3. Preparation of the Tri-Block Copolymer #17

(9) In a manner similar to example 1 an acylated PLCA-PEG-PLCA tri-block was synthesized having a PLCA/PEG ratio of 1.8 and a caprolactone/lactide weight ratio 9/1. The difference with example 1 was: different PLCA/PEG ratio.

Example 4 Fully C2-Modified PLCA-PEG-PLCA (#21)

(10) Polyethyleneglycol (PEG 1500, 10 g, 6.7 mmol) and ca. 110 ml toluene were charged into a 250 ml three-neck round-bottom flask equipped with a magnetic stirring bar. Using a Dean-Stark device with a condenser on top, 45 ml of toluene was distilled off to remove water (from PEG) azeotropically by heating at 150 C. at atmospheric pressure under nitrogen.

(11) After cooling down the solution to ca. 80-100 C., L-lactide (2 g, 14 mmol) and caprolactone (18 g, 158 mmol) were added. 40 ml of toluene was distilled off to dry the monomers by heating at 150 C. at atmospheric pressure. Ca. 25 ml of dry toluene was left in the flask for the polymerization.

(12) After cooling down the mixture to ca. 80-100 C., tin(II) 2-ethylhexanoate (0.2 ml) was added through one of the necks.

(13) Polymerization was carried out at 120 C. for 1 day under nitrogen atmosphere.

(14) After cooling down to room temperature, ca. 60 ml dichloromethane and 4.7 ml triethylamine (34 mmol) were added. Subsequently, 2 ml acetyl chloride (28 mmol) was added slowly to the stirred solution, which was cooled with an ice bath. The acylation reaction was continued for a few hours after which dichloromethane was removed by rotavap, and ethyl acetate (ca. 100 ml) was added to the residue. Triethylamine salt was removed by (paper) filtration and the polymer, which was dissolved in the clear filtrate was precipitated by addition of a (1:1) mixture of hexane and diethyl ether. At ca. 20 C. (in freezer) the polymer product separated as a waxy solid from which non-solvents could be decanted easily. Finally, the precipitated polymer was dried in vacuo. Yield: ca. 27 g.

(15) The PLCA/PEG block ratio is around 2; caprolactone/lactide weight ratio 9/1. Degree of modification (from NMR) is ca. 2.

Example 5. Synthesis Comparative Tri-Block Copolymer U5

(16) This is the polymer that is similar to the one mentioned in the examples of US2007/0265356 (same block ratio, but different caplac).

(17) Polyethyleneglycol (PEG 1500, 12.5 g, 8.3 mmol) and ca. 125 ml toluene were charged into a 250 ml three-neck round-bottom flask equipped with a magnetic stirring bar. Using a Dean-Stark device with a condenser on top, 60 ml of toluene was distilled off to remove water (from PEG) azeotropically by heating at 150 C. at atmospheric pressure under nitrogen.

(18) After cooling down the solution to ca. 80-100 C., L-lactide (3 g, 21 mmol) and caprolactone (12 g, 105 mmol) were added. 40 ml of toluene was distilled off to dry the monomers by heating at 150 C. at atmospheric pressure. Ca. 25 ml of dry toluene was left in the flask for the polymerization.

(19) After cooling down the mixture to ca. 80-100 C., tin(II) 2-ethylhexanoate (0.2 ml) was added through one of the necks.

(20) Polymerization was carried out at 120 C. for 1 day under nitrogen atmosphere.

(21) After cooling down to room temperature, ca. 50 ml dichloromethane and 2.5 ml triethylamine (18 mmol) were added. Subsequently, 2.1 ml hexanoyl chloride (15 mmol) was added slowly to the stirred solution, which was cooled with an ice bath. The acylation reaction was continued for a few hours after which dichloromethane was removed by rotavap, and ethyl acetate (ca. 100 ml) was added to the residue. Triethylamine salt was removed by (paper) filtration and the polymer, which was dissolved in the clear filtrate was precipitated by addition of a (1:1) mixture of hexane and diethyl ether. At ca. 20 C. (in freezer) the polymer product separated as a waxy solid from which non-solvents could be decanted easily. Finally, the precipitated polymer was dried in vacuo. Yield: ca. 27 g.

(22) The PLCA/PEG block ratio of the comparative U5 block copolymer is around 1.2; caprolactone/lactide weight ratio 4/1. Degree of modification (from NMR) is ca. 1.7

(23) Properties of the polymer are listed in table 1 below.

(24) TABLE-US-00001 TABLE 1 Characteristics of tri-block copolymers #33, #27, #21, #17 and U5 Block ratio Weight ratio of Mn as Degree of Tri-block (PLCA/PEG -caprolactone to determined End- acylation (degree of copolymer ratio) lactide (cap/lac ratio) using .sup.1H NMR group modification, DM) (%) Comparative U5 1.2 4/1 3300 Hex- 75 polymer anoyl (C6) PLCA/PEG2.2- 2.2 4/1 5300 Acetyl 100 CL/LA 4/1-C2 (C2) (#17) PLCA/PEG2.0- 2.0 9/1 4700 Acetyl 100 CL/LA 9/1-C2 (C2) (#21) PLCA/PEG1.8- 1.8 9/1 4200 Acetyl 100 CL/LA 9/1-C2 (C2) (#27) PLCA/PEG2.2 2.2 9/1 5300 Acetyl 100 CL/LA 9/1-C2 (C2) (#33)

Example 6 Release of Model Drugs in PBS

(25) This example illustrates the release profile of water soluble and hydrophobic model drugs from tri-block copolymer solutions of the present invention.

(26) The following water soluble model proteins were used:

(27) (1) IgG antibody (Sigma); this is a protein of around 150 kDa and has a solubility in water is above 10000 g/ml.

(28) (2) bovine serum albumin (Sigma); this is a protein of around 40 kDa and has a solubility in water of above 10000 g/ml.

(29) (3) lysozyme (Sigma); this is a protein of around 16 kDa and has a solubility in water of above 10000 g/ml.

(30) The following small molecules were used:

(31) (1) celecoxib (LC Laboratories); this is a non-steroidal anti-inflammatory drug of 386 g/mol and has a solubility in water below 1 ug/ml

(32) (2) triamcinolone acetonide (Sigma); this is a corticosteroid and has a solubility in water below 20 ug/ml

(33) (3) iobitridol (Guerbet); this is a contrast agent of around 700 g/moland has a solubility in water above 5000 ug/ml.

(34) The following peptide was used:

(35) (1) Cyclosporin A (Sigma); this is a non-natural cyclic peptideof around 1000 g/mol amd has a solubility is water below 5 ug/ml

(36) Polymer solutions were prepared with the tri-block copolymers described in Table 1. The tri-block copolymers were weighted in 7-ml glass vial and melted at 50 C. using a water bath. Protein stock solutions in PBS (phosphate buffer saline; B. Braun, 10 mM, pH=7.4) were added to the molten polymer to achieve the required polymer content in the solutions containing 1% (w/v) protein. The mixtures of protein, tri-block copolymer and PBS were vortexed for 1 min and placed on a rotating wheel at 4 C. overnight.

(37) For the hydrophobic drugs, a mixture drug/polymer was first prepared in ethylaacetate. Ethylacetate was evaporated and PBS was added.

(38) With a syringe, 300 l of the solutions were placed in glass vials (8.240 mm). The glass vials were incubated for 30 min at the required release temperature to allow gelation. Pre-heated PBS was added on top of the gel. At regular time intervals, the buffer was removed, followed by the addition of fresh PBS. The removed buffer samples were analyzed for their protein content by either UPLC (for BSA and lysozyme) or bicinchoninic acid assay (BCA) protein assay (for IgG).

(39) IgG was analysed by BCA protein assay: IgG release samples (25 L) were pipetted into a 96-microwells plate and 200 L of working reagent (BCA reagent A: BCA reagent B, 50/1 v/v) was added. After incubation of the plates for 30 minutes at 37 C., the absorbance was measured at 550 nm with a Novapath Microplate Reader (Bio-Rad). Standard protein solutions (concentration range: 0.01-1 mg/mL) were prepared to generate a calibration line.

(40) BSA was analysed by UPLC: BSA release samples were analyzed using a Waters Acquity Ultra Performance LC system equipped with an Acquity BEH300 C18 1.7 m column (5 cm), a binary solvent manager, a sample manager with column oven at 50 C. and an Acquity TUV detector (detection wavelength: 210 nm). After injection of 5 L release sample, a gradient was run from 80% A (H2O/ACN 95/5 (% v/v) containing 0.1% TFA) to 60% B (100% ACN with 0.1% TFA) in 2 minutes at a flow rate of 0.25 mL/min. Standard protein solutions (concentration range: 0.01-1 mg/mL) were used to obtain a calibration line.

(41) Lysozyme was analysed by UPLC: lysozyme release samples were analyzed using a Waters Acquity Ultra Performance LC system equipped with an Acquity BEH300 C18 1.7 m column (5 cm), a binary solvent manager, a sample manager with column oven at 50 C. and an Acquity TUV detector (detection wavelength: 280 nm). After injection of 5 L release sample, a gradient was run from 75% A (H2O/ACN 95/5 (% v/v) containing 0.1% TFA) to 40% B (100% ACN with 0.1% TFA) in 4 minutes at a flow rate of 0.25 mL/min. Standard protein solutions (concentration range: 0.01-1 mg/mL) were used to obtain a calibration line.

(42) Celecoxib was analysed by UPLC: Celecoxib concentration in the buffer was measured by a Waters Acquity) Ultra Performance LC system (UPLC, Waters, Milford Mass., USA) equipped with an Acquity BEH C18 1.7 m column (2.1100 mm), a binary solvent manager, a sample manager with column oven at 50 C. and an Acquity TUV Detector (detection wavelength: 254 nm). After injection of 7.5 l release sample, a gradient was run from 100% A (H2O/ACN 95/5% (v/v) containing 0.1% TFA) to 100% B (MeOH/ACN/H2O 45/45/10% (v/v) containing 0.1% TFA) in 2 minutes and kept at 100% B from 10 min at a flow rate of 0.08 ml/min. The run time was 16 min. A calibration curve was obtained after injection of standard celecoxib solutions in DMSO (0.5-100 ug/ml). The chromatograms were analused using Empower Software Version 1154 (Waters, Milford Mass., USA). CLB concentration in the release sample was in the range 5-20 g/ml.

(43) Triamcinolone acetonide (TA) was analysed by UPLC: TA concentration in the buffer was measured by a Waters Acquity Ultra Performance LC system (UPLC, Waters, Milford Mass., USA) equipped with an Acquity BEH C18 1.7 m column (2.150 mm), a binary solvent manager, a sample manager with column oven at 50 C. and an Acquity TUV Detector (detection wavelength: 240 nm). After injection of 7.5 l release sample, a gradient was run from 50% A (H2O/ACN 95/5% (v/v) containing 0.1% TFA) to 80% B (ACN/H2O 95/10% (v/v) containing 0.1% TFA) in 1.5 minutes at a flow rate of 0.25 ml/min. The run time was 3.2 min. A calibration curve was obtained after injection of standard TA solutions in DMSO (0.5-100 ug/ml). The chromatograms were analused using Empower Software Version 1154 (Waters, Milford Mass., USA). TA concentration in the release sample was in the range 5-20 g/ml.

(44) Iobitriodol (ITD) was analysed by UPLC: ITD concentration in the buffer was measured by a Waters Acquity) Ultra Performance LC system (UPLC, Waters, Milford Mass., USA) equipped with an Acquity HSS 1.7 m column (2.150 mm), a binary solvent manager, a sample manager with column oven at 50 C. and an Acquity TUV Detector (detection wavelength: 204 nm). After injection of 7.5 l release sample, a gradient was run from 100% A (H2O 2O containing 0.1% TFA) to 48% B (IPA containing 0.1% TFA) in 0.9 minutes at a flow rate of 0.25 ml/min. The run time was 2 min. A calibration curve was obtained after injection of standard ITD solutions in PBS (0.5-100 ug/ml). The chromatograms were analysed using Empower Software Version 1154 (Waters, Milford Mass., USA). ITD concentration in the release sample was in the range 5-20 g/ml.

(45) Cyclosporin A (CyA) was analysed by UPLC: CyA concentration in the buffer was measured by a Waters Acquity) Ultra Performance LC system (UPLC, Waters, Milford Mass., USA) equipped with an Acquity BEH C4 1.7 m column (2.150 mm), a binary solvent manager, a sample manager with column oven at 50 C. and an Acquity TUV Detector (detection wavelength: 220 nm). After injection of 7.5 l release sample, a gradient was run from 100% A (H2O/ACN 95/5% (v/v) containing 0.1% TFA) to 100% B (ACN/H2O 95/10% (v/v) containing 0.1% TFA) in 1 minutes at a flow rate of 0.25 ml/min. The run time was 3.6 min. A calibration curve was obtained after injection of standard CyA solutions in DMSO (0.5-100 ug/ml). The chromatograms were analysed using Empower Software Version 1154 (Waters, Milford Mass., USA). CyA concentration in the release sample was in the range 5-20 g/ml.

(46) The results of this example are shown in FIGS. 1 to 4.

Example 7 Synthesis of B-A-B Block

(47) The syntheses of the compositions were conducted as per Example 4.

(48) TABLE-US-00002 TABLE 2 Characteristics of tri-block copolymers Weight ratio of Block ratio caprolactone to (PLCA/PEG lactide Degree of ID Tri-block copolymer ratio) (cap/lac ratio) End group acylation/% A PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.1.8(C.sub.2).sub.2.0 1.8 9/1 Acetyl C2 100 B PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.0(C.sub.2).sub.2.0 2.0 9/1 Acetyl C2 100 C PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.2(C.sub.2).sub.2.0 2.2 9/1 Acetyl C2 100 D PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.4(C.sub.2).sub.2.0 2.4 9/1 Acetyl C2 100 E PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.6(C.sub.2).sub.2.0 2.6 9/1 Acetyl C2 100 F PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.8(C.sub.2).sub.2.0 2.8 9/1 Acetyl C2 100 G PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.0(C.sub.2).sub.0 2.0 9/1 0 H PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.0(C.sub.2).sub.1 2.0 9/1 Acetyl C2 50 I PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.1..sub.2(C.sub.2).sub.2.0 1.2 9/1 Acetyl C2 100 J PEG.sub.1500(CAP.sub.80%/LAC.sub.20%).sub.1..sub.2(C.sub.6).sub.2.0 1.2 4/1 Hexanoyl 100 C6 K PEG.sub.2000(CAP.sub.80%/LAC.sub.20%).sub.1..sub.2(C.sub.6).sub.2.0 1.2 4/1 Hexanoyl 100 C6

Example 8 Degradation of Hydrogel Compositions with Varying PEG/PLCA Ratios

(49) This example shows the degradation profiles for compositions according to the invention. Typically, gels were cast from 20 wt % polymers solution with 3 wt % lysozyme in small vials and allowed to gel at 37 C. PBS buffer was added and replaced periodically. The wet gels were weighed and the amount of lysozyme released was determined by the BOA assay and/or UPLC (BOA and UPLC were conducted as per example 6). The enzymatic activity of lysozyme was determined by the hydrolysis of the outer cell membrane of Micrococcus Lysodeikticus.

(50) TABLE-US-00003 TABLE 3 Degradation of the B-A-B tri-block copolymers Unloaded gels Protein-loaded gels.sup.1 ID Structure at 20 wt % at 25 wt % BSA Lysozyme I PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.1.2(C.sub.2).sub.2.0 COMPARATIVE EXAMPLE: does not form a gel at 37 C. A(#27) PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.1.8(C.sub.2).sub.2.0 8 days 14 days 6 days 6 days B (#21) PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.0(C.sub.2).sub.2.0 27 days 31 days 21 days 19 days C (#33) PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.2(C.sub.2).sub.2.0 31 days 34 days 12 days 12 days D PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.4(C.sub.2).sub.2.0 34 days 34 days 11 days 12 days E PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.6(C.sub.2).sub.2.0 250 days 250 days 12 days 12 days F PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.8(C.sub.2).sub.2.0 COMPARATIVE EXAMPLE: does not dissolve in PBS

Example 8b

(51) TABLE-US-00004 TABLE 4 Different end-groups Weight ratio Block ratio of caprolactone (PLCA/PEG to lactide Degree of ID Tri-block copolymer ratio) (cap/lac ratio) End group acylation/% Z1 PEG1500(CAP90%/LAC10%)1.8(C3)2.0 1.8 9/1 C3 100 Z2 PEG1500(CAP90%/LAC10%)2(C3)2.0 2 9/1 C3 100 Z3 PEG1500(CAP90%/LAC10%)1.8(C4)2.0 1.8 9/1 C4 100 A PEG1500(CAP90%/LAC10%)1.8(C2)2.0 1.8 9/1 Acetyl 100 C2 B PEG1500(CAP90%/LAC10%)2.0(C2)2.0 2.0 9/1 Acetyl 100 C2

(52) Tri-block copolymers Z1-Z3 were synthesised in a manner similar to examples 2 and 4.

(53) 20% solutions of the tri-block copolymers of Table 4 in PBS were prepared and placed in an oven at 37 C. for thermogelations. The forming of the gel was checked after 1 hour and gels were formed for compositions comprising tri-block copolymers comprising A, B, Z1, Z2 and Z3. After longer periods of time (several hours) also Z3 showed phase separation and after even longer periods Z2 and Z1 also showed phase separation. Phase separation was not observed for compositions comprising tri-block copolymers A and B.

Example 9 Release of Model Protein Vs PCLA/PEG Ratio

(54) This example illustrates the release profile of water soluble model proteins from tri-block copolymer solutions of the present invention.

(55) The method for detecting release of the drug from the compositions was followed as per example 2. Burst release was calculated by plotting the cumulative release versus t.sup.0.5. The degradation experiments were conducted as per example 2. The results are presented in Table 4.

(56) TABLE-US-00005 TABLE 5 Characteristics of tri-block copolymers with varying weight ratio -caprolactone to lactide Weight ratio Release of -caprolactone Degradation water soluble Burst ID Structure to lactide Time model protein release M PEG.sub.1500(CAP.sub.100%/LAC.sub.0%).sub.1.8(C.sub.2).sub.2.0 1/0 >140 days ~25 days 1.8% -caprolactone N PEG.sub.1500(CAP.sub.95%/LAC.sub.5%).sub.2.0(C.sub.2).sub.2.0 19/1 >50 days ~25 days 1.6% B(21) PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.0(C.sub.2).sub.2.0 9/1 ~42 days ~30 days 1.0% O PEG.sub.1500(CAP.sub.80%/LAC.sub.20%).sub.2.0(C.sub.2).sub.2.0 4/1 ~23 days ~16 days 4.9% P PEG.sub.1500(CAP.sub.50%/LAC.sub.50%).sub.2.0(C.sub.2).sub.2.0 1/1 ~22 days ~9 days 5.0%

Example 10 Rheology of Compositions with Different End Caps

(57) Composition B (#21) from experiment 9 (which had the minimum burst release profile) was compared by rheological analysis, to the compositions with the same PLCA/PEG ratio but with different degrees of end acylation.

(58) Rheological characterization of the blends was done with a AR-G2 rheometer (TA Instruments, Etten-Leur, The Netherlands) equipped with a 1 steel cone geometry of 20 mm diameter and solvent trap. Polymer blend solutions of 20% (w/w) were prepared in PBS pH 7.4 at 4 C. 300 l of the solutions were placed in glass vial (8.240 mm) and incubated for around 3 hours at 37 C. to enable gelation and stabilization. Using a spatula, approximately 70 mg of the sample was placed between the pre-heated (37 C.) plates of the rheometer. Rheological gel characteristics were monitored by oscillatory time sweep experiments. During time sweep experiments G(shear storage modulus) and G (loss modulus) were measured for a period of 5 min. Also temperature sweep experiments were performed on the polymer solutions. Therefore the plates of the rheometer were pre-cooled at 4 C. Temperature increase was 1 C./min. When G/G (=tan )=1, the sample is considered as a gel in a rheological point of view. All experiments were performed at constant strain (1%) and frequency (1 Hz).

(59) The results are shown in Table 6

(60) TABLE-US-00006 TABLE 6 Gel characteristics for compositions B and F Gel at Degree of Composition 37 C. acylation/% B 21 YES 100 F PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.0(C.sub.2).sub.1 YES 50

Example 11 Degradation of Compositions of Tri-Block Copolymers of the Invention with Different Degrees of Acylation

(61) This example illustrates the degradation profile for tri-block copolymers of the invention compared to a model compound that is not acylated. In this example 300 mg of a formulation containing 20% tri-block copolymer and 3 wt % lysozyme was prepared. The procedure was followed according to example 8.

(62) The results are shown in FIG. 5 and summarized in Table 7.

(63) TABLE-US-00007 TABLE 7 Degradation times for tri-block copolymers Degradation Degree of ID Polymer composition time acylation/% F PEG1500(CAP90%/LAC10%).sub.2.0(C2).sub.0 14 days 0 G PEG1500(CAP90%/LAC10%).sub.2.0(C2).sub.1 14 days 50 B PEG1500(CAP90%/LAC10%).sub.2.0(C2).sub.2 42 days 100

Example 12 Release of Lysozyme

(64) This example shows the release profile of lysozyme which was used as a model water soluble protein, from compositions of the invention compared to a composition comprising a tri-block copolymer that is not acylated.

(65) The procedure was followed as per example 6.

(66) The results are shown in FIG. 6 and summarized in Table 8.

(67) TABLE-US-00008 TABLE 8 Comparison of controlled release profile for tri-block copolymers with different degrees of acylation. Degree of Controlled ID Composition acylation/% burst release* F PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.0(C.sub.2).sub.0 0 G PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.0(C.sub.2).sub.1 50 + B PEG.sub.1500(CAP.sub.90%/LAC.sub.10%).sub.2.0(C.sub.2).sub.2. 100 ++ very large Durst release large burst release + small burst release ++ very small burst release

Example 13 Mathematical Analysis of Release Profiles

(68) The data from the release experiment (example 12) was plotted against the square root of the time of release. The data was fitted using linear regression analysis and the burst release could be determined by extrapolating the lines of best fit to the y axis. The point where the extrapolated line crosses the y axis gives the value for percentage burst release. The results are shown in FIG. 7.

Example 14 Synthesis of Iodine Functionalized Tri-Block Copolymers

(69) This example details a method to covalently bond a radiopaque atom to a tri-block copolymer. The structure of a tri-block copolymer which is acylated is given below in formula (1).

(70) ##STR00001##

(71) The synthesis of tri-block copolymers wherein at least part of the hydroxyl end-groups of the tri-block copolymer are covalently bound to a compound containing radiopaque atoms was done according to the method previously described in US7740877, examples 1, 2 and 3. The synthesis led to tri-block copolymers (Table 9) with triiodobenzoyl groups (1.8 groups per chain) as confirmed by 1H NMR.

(72) TABLE-US-00009 TABLE 9 Tri-block copolymers used. Number average mol. Caprolactone/ Degree of weight ID ID PLCA/PEG lactide acylation (%) End-group PEG P I- 1.8 9/1 90% Triiodobenzoyl 1500 (#81) Gell #1 Q I- 1.0 9/1 90% Triiodobenzoyl 1500 (#29) Gell #2

(73) In the following examples different compositions comprising P and Q were made and the compositions are summarized in Table 10.

(74) TABLE-US-00010 TABLE 10 Overview of compositions used in examples 15-19 Tri-block copolymer (wt %) Example P Q 17 15 100 16 25 75 30 70 50 50 70 30 17 a 30 70 b 30 70 18 a 100 b 25 75 19 a 25 75 b 50 50

Example 15 Loading of a Tri-Block Copolymer Composition with Contrast Agent Hexabrix

(75) In order to achieve prolonged release of a pharmaceutically active ingredient, retention in the treatment location (for example a joint) must be maintained for a minimum of 4 weeks. To determine this retention time the tri-block copolymer composition must be visualized. This example shows how loading a tri-block copolymer with a conventional contrast agent does not achieve this goal.

(76) Polymer #17 (see Table 1) was loaded with 15% Hexabrix (Guerbet). Hexabrix contrast agent is a low osmolar ionic dimer. Each milliliter contains 393 mg of ioxaglate meglumine, 196 mg of ioxaglate sodium and 0.10 mg of edetate calcium disodium as a stabilizer. The solution contains 3.48 mg (0.15 vmEq) sodium in each milliliter and provides 32% (320 mg/mL) organically bound iodine.

(77) The commercially available Hexabrix solution was diluted with phosphate buffer (50 mM phosphate, pH 7.4, 0.07 mM NaCl) at a ratio 15/85 v/v. This diluted Hexabrix solution was used to dissolve the polymer, leading to the preparation of a formulation containing 25 wt % polymer with respect to buffer and Hexabrix. Each gram of formulation contained 44 mg ioxaglate meglumine.

(78) The loaded tri-block copolymer composition was injected into a chicken knee and visualized using microCT (Skyscan model 1076, Skyscan, Kontich, Belgium) Scans were performed using the following scanner settings: isotropic voxelsize of 35 mm, at a voltage of 55 kV, a current of 170 mA, field of view of 35 mm, and a 0.5 mm aluminium filter, over 198 with a 0.8 rotation step. 1 hour after injection the visibility of the gel was diminished and a further 1 day later the composition could not be visualized. Make table T=0-1-24 hrs showing 100%-50%-0% visualization as relative grey value

(79) The results are also indicated in Table 11.

(80) TABLE-US-00011 TABLE 11 Image-retention of Polymer #17 (25% w/w) loaded with 15% Hexabrix Time point in hours Relative grey surface area T = 0 100% T = 1 50% T = 24 0%

Example 16 Radiopaque Tri-Block Copolymer Compositions

(81) This example shows that a tri-block copolymer wherein at least part of the hydroxyl end-groups of the tri-block copolymer are covalently bound to a compound containing radiopaque atoms can be visualized by microCT. (Skyscan model 1076, Skyscan, Kontich, Belgium) Scans were performed using the following scanner settings: isotropic voxelsize of 35 mm, at a voltage of 55 kV, a current of 170 mA, field of view of 35 mm, and a 0.5 mm aluminium filter, over 198 with a 0.8 rotation step. Scan time was 10 minutes. In this example different compositions were prepared comprising different weight percentages of tri-block copolymers (Table 10, example 16) wherein the hydroxyl end-groups of the tri-block copolymer are at least partially acylated with an acetyl group or wherein at least part of the hydroxyl end-groups of the tri-block copolymer are covalently bound to a compound containing radiopaque atoms.

(82) The compositions (Table 10, examples 16) were prepared as follows. Both polymers P and #17 were separately dissolved in ethylacetate at a concentration of 500 mg/mL. The solutions were mixed at the desired ratios and the mixtures were transferred into petri dishes. The solvent was removed under nitrogen flow for 48 hours. To 500 mg polymer blend, phosphate buffer (50 mM, 0.07 mM NaCl, 0.02% NaN3, pH 7.4) was added to yield solutions at 25 wt %. The compositions were measured in glass vials and the X-ray intensity plotted on a graph (FIG. 8). The compositions were measured in glass vials and the X-ray intensity plotted on a graph (FIG. 8).

Example 17 CT Imaging in Radiopaque Tri-Block Copolymer Compositions

(83) This example shows the CT visualization times for different mixtures of tri-block copolymers.

(84) The mixtures of copolymers (Table 10, example 17a and 17b) were prepared, injected into the knee of a rat cadaver and visualized by microCT (Skyscan model 1076, Skyscan, Kontich, Belgium). Scans were performed using the following scanner settings: isotropic voxelsize of 35 mm, at a voltage of 55 kV, a current of 170 mA, field of view of 35 mm, and a 0.5 mm aluminium filter, over 198 with a 0.8 rotation step. The results are shown in Table 14.

(85) The composition 17a was visible for only a few days whereas composition 17b was visible for 3 weeks (FIG. 11).

Example 18 Release Profiles of Tri-Block Copolymers Wherein the End Groups are Acylated or are Radiopaque Substituted Acyl Groups

(86) This example illustrates the release profile of a pharmaceutically active ingredient from tri-block copolymer compositions and also the degradation profiles for tri-block copolymer compositions comprising a pharmaceutically active ingredient.

(87) OAc-Gell (#17) and a mixture of a I-Gell#1 (P) and OAc-Gell (#17) were prepared at 25 wt % in 50 mM phosphate buffer at pH 7.4, 0.42% NaCl and 0.05% NaN3. The loading of Celecoxib was 1.25 mg/mL. Release experiments were performed as per example 6 and at 37 C. in PBS buffer containing 0.2% Tween 80. Error bars represent standard error of the mean (n=6). The results are shown in FIG. 9 and summarized in Table 12.

(88) TABLE-US-00012 TABLE 12 Overview release time and erosion profile for examples 18a and 18b. Example Release time Erosion time (solid content) 18a 28 28 18b 35 35

(89) Wet weight=weight of the depots, as measured by decanting the PBS buffer and weighing the remaining gel, after which the original vial-weight is subtracted

(90) Dry weight=weight of the freeze dried depots after decanting of the buffer (i.e. polymer weight) and subtracting the weight of the vial

(91) Solid content=polymer concentration in the depots (dry weight/wet weight)

Example 19 Degradation Profile for Tri-Block Copolymer Compositions

(92) This example shows the in vivo degradation profile for compositions of tri-block copolymers wherein at least part of the hydroxyl end-groups of the tri-block copolymer are covalently bound to a compound containing radiopaque atoms.

(93) Compositions (Table 10, example 19a and 19b) were prepared were prepared by dissolving 25 wt % tri-block copolymer (comprising 75% P and 25% #17, or 50% P and 50% #17) in 75% PBS buffer, pH 7.4. 50 mM, 0.15% NaCl of the composition was injected subcutaneously into rats knees. The injected volume of 19b was then visualized and measured using 3D micro-CT imaging (Skyscan model 1076, Skyscan, Kontich, Belgium). The rats were anesthetized using Isoflurane and then placed in a custom made scanner bed, fixing the hind limb in an extended position. Scans were performed using the following scanner settings: isotropic voxel size of 35 mm, at a voltage of 55 kV, a current of 170 mA, field of view of 35 mm, and a 0.5 mm aluminium filter, over 198 with a 0.8 rotation step. The rat knees and subcutaneous depots were scanned on days 0, 1, 4 and 8 and after that weekly until the gel was no longer visible. The scan time was 16 minutes and a frame averaging of 3 was used. Images obtained using the uCT scanner were reconstructed using Skyskan analysis software. The datasets were segmented using a fixed attenuation threshold between air and subchondral bone. Subsequently, regions of interest were drawn around the patellar cartilage and attenuation and volume were calculated. The injected volume of 19a (non-radiopaque), was monitored through the skin with a measuring caliper. The results are shown in FIG. 10 and summarized in Table 13 and 14.

(94) TABLE-US-00013 TABLE 13 Overview of erosion monitoring for examples 19a and 19b. 19a 19b Day Observations Surface area/% Observations Surface area/% 0 Distinct oval 100 Higher 100 shape, sharp intensity edges than for 19a 4 Sharp edges 80 70 maintained 8 Less sharp 40 65 edges on interior side 14 Fractured 10 75 shape - small light spots in dark area 21 Not visible 50 40 Still visible 40

(95) Surface area=relative area with respect to surface area at t=0 (set to 100%)

(96) TABLE-US-00014 TABLE 14 Overview of compositions and microCT characteristics for examples 15-19 Tri-block copolymer Duration of (wt %) visibility Active Example P Q 17 in micro CT ingredient 15 100 <24 hours Hexabrix 16 a 25 75 All visible in micro 30 70 CT but duration 50 50 of visibility 70 30 not measured 17 a 30 70 <7 days b 30 70 ~4 weeks 18 a 100 Celecoxib b 25 75 Celecoxib 19 a 25 75 ~40 days b 50 50 ~77 days

(97) Discussion and Conclusion

BRIEF DESCRIPTION OF THE FIGURES

(98) FIG. 1 shows the release of both hydrophobic-type (celecoxib (CLB), triamcinolone (TA), cyclosporine A (cya) and water soluble (iobitrodol (ITD) and lysozyme (Lys)) active ingredients from 25% w/w U5 (low PLCA/PEG ratio) based on PBS and active ingredient at 37 C. The comparative U5 polymer that phase separates at 37 C.

(99) In FIG. 1, the following legend is used:

(100) Release from U5: Lysozyme (diamonds), ibitridol (circles), Triamcinolone acetonide (squares), Cyclopsorin A (open squares), celecoxib (open triangles)

(101) FIG. 2 shows the release of IgG, BSA and lysozyme from a solution of 20% w/w tri-block copolymer #17 based on the amount of solvent and pharmaceutically active ingredient present in the composition at 33 C. Data are shown as averagestandard deviation (n=3).

(102) FIG. 3 shows the release of IgG, BSA and lysozyme from a solution of 20% w/w tri-block copolymer #33 based on the amount of solvent and pharmaceutically active ingredient present in the composition at 33 C.

(103) FIG. 4 shows the release of model proteins lysozyme, BSA and IgG from tri-block copolymers #27 and from #21 at 37 C. at a concentration of 20% w/w tri-block copolymer based on the amount of solvent and pharmaceutically active ingredient present in the composition. Data are shown as averagestandard deviation (n=3).

(104) FIG. 5 shows the degradation profile of the relative wet gel as a function of time for formulations of 20 wt % of polymer and 3 wt % of lysozyme. Measurements were made on 4 samples (n=4)

(105) In FIG. 5 the compositions used are

(106) G PEG1500(Cap90%/Lac10%)2.0 (triangle)

(107) H PEG1500(Cap90%/Lac10%)2.0(C2)1 (diamond)

(108) B (#21): PEG1500(Cap90%/Lac10%)2.0(C2).sub.2 (square)

(109) FIG. 6 shows the cumulative release profile of lysozyme as a function of time for formulations of 20 wt % of tri-block copolymer and 3 wt % of lysozyme.

(110) In FIG. 6 the compositions used are

(111) G PEG1500(Cap90%/Lac10%)2.0 (triangle)

(112) H PEG1500(Cap90%/Lac10%)2.0(C2)1 (diamond)

(113) B (#21): PEG1500(Cap90%/Lac10%)2.0(C2).sub.2 (square)

(114) FIG. 7 shows the cumulative release profiles of lysozyme from tri-block copolymer compositions as a function of the square root of time. The formulations comprised 20 wt % of tri-block copolymer and 3 wt % of lysozyme. Lines represent linear fits. The fitted curves correspond to y=m+c where c is the percentage burst release. Error bars show the standard deviation with n=4.

(115) In FIG. 7 the compositions used are

(116) G PEG1500(Cap90%/Lac10%)2.0 (triangle)

(117) H PEG1500(Cap90%/Lac10%)2.0(C2)1 (diamond)

(118) B (#21): PEG1500(Cap90%/Lac10%)2.0(C2).sub.2 (square)

(119) FIG. 8 shows the x-ray intensities as measured by the microCT, for compositions shown in Table 10, example 16, whereby the percentage of iodine bound polymer P(#81) is given on the x-axis. The line is added to guide the eye.

(120) FIG. 9 shows the release profile and gel erosion profiles for example 18. The filled squares indicate composition 18a (OAc-Gell only loaded with Celecoxib) and the empty triangles indicate composition 18b (OAc-Gell/I-Gell#1-#81). The compositions contained 25 wt % in 50 mM phosphate buffer, pH 7.4, 0.42% NaCl and 0.05% NaN3. The loading of Celecoxib was 1.25 mg/mL

(121) FIG. 10 shows the erosion profile for compositions 19a (dotted line) and 19b (Solid line). Images of the subcutaneous injected I-Gell depots in rats were taken in-vivo using micro-CT on days 0-1-4-8-14-21-28-35-42-49-56-63-70-77 and the volume of the gel was determined using a caliper as done routinely for volume-estimations of subcutaneous tumor lumps.

(122) FIG. 11 shows the degradation profile for composition 17b as determined by visualization using microCT.

(123) As can be seen from FIG. 1, the tri-block copolymer with a low PLCA/PEG ratio of 1.2 maybe be employed to obtain long-term controlled release of hydrophobic molecules (celecoxib, triamcinolone acetonide, cyclosporin A) without burst; whereas that polymer is not suitable to prevent burst release of water soluble molecules (lysozyme, iobitridol) and as such not suitable for use in the composition of the present invention.

(124) As can be seen from FIG. 2, the tri-block copolymer #17 may be employed in the composition of the invention to get controlled release without too much burst release of water soluble model proteins IgG; BSA and lysozyme.

(125) As can be seen from FIG. 3, the tri-block copolymer #33 may be employed to get a controlled release without too much burst release of the water soluble pharmaceutically active ingredients (in this case Lysozyme, BSA and IgG).

(126) As can be seen from FIG. 4, the tri-block copolymers #21 and #27 are suitable for the controlled release of the tested model proteins: lysozyme and BSA.

(127) The results in Table 3 show that at a block ratio of lower than 1.4, the tri-block co-polymer does not form a gel whereas at a block ratio of above 2.6, the tri-block co-polymer does not dissolve in the buffer solution.

(128) The gel behaviour for the compositions shown in Table 4 indicate that preferably for release of water soluble proteins, the hydroxyl end-groups of the tri-block copolymer in the composition of the invention are at least partially acylated with an optionally substituted acetyl, an optionally substituted propionyl, an optionally substituted butyl, or an optionally substituted benzoyl group. Preferably, the hydroxyl end-groups of the tri-block copolymer in the composition of the invention are at least partially acylated with an optionally substituted acetyl or an optionally substituted propionyl. More preferably, the hydroxyl end-groups of the tri-block copolymer in the composition of the invention are at least partially acylated with an optionally substituted acetyl or an optionally substituted propionyl, most preferably with an optionally substituted acetyl.

(129) As can be seen from Table 5 longer degradation and release are obtained for tri-block copolymers having a weight ratio of -caprolactone to lactide of at least 9/1. Furthermore it can be seen that a tri-block copolymer having a weight ratio of -caprolactone to lactide of at least 1/1 shows low burst release.

(130) Table 6 shows that the at least partially acylated tri-block copolymers of the invention can form a gel.

(131) As shown in FIG. 5 and Table 7, the composition having a degree of acylation of the end-groups of at least 75% for example of 100% have a significantly longer degradation time (42 days) than those compositions having a lower degree of acylation.

(132) As can be seen from FIG. 6, the compositions comprising a tri-block copolymer that is at least partially acylated (degree of acylation 50%) or fully acylated (degree of acylation 100%) provide gels with a lower burst release, when compared to compositions comprising a tri-block copolymer that is not acylated. This example clearly demonstrates the surprising improvement in terms of release profile of a water soluble protein from a composition according to the invention comprising an at least partially acylated tri-block copolymer e.g. having a degree of acylation of at least 75% in comparison with the composition comprising a tri-block copolymer lacking acylation.

(133) As can clearly be seen from FIG. 7, compositions comprising at least partially acylated (e.g. degree of acylation 50%) or fully acylated end-groups (e.g. degree of acylation 100% tri-block co polymers) have a burst release of less than 2%, which is surprising and significant improvement for use in controlled release of especially water soluble compounds, for example water soluble proteins. The composition comprising non-acylated caprolactone/lactide showed a burst release of 24% with the same water soluble compounds. which renders this composition useless, and even unsafe, in clinical practise for most protein-based drug therapies.

(134) The conclusion that can be drawn from the mathematical analysis (as shown in FIG. 7) of the release profile is that compositions of the invention comprising tri-block copolymers that are at least partially acylated give low or very low burst release compared to compositions comprising tri-block copolymers that are not acylated.

(135) It is also observed, that when using triblock copolymers in the composition of the invention, wherein the number average molecular weight of the A-block is 1000 Da, we obtain gels at 37 C. that immediately phase separate. When using triblock copolymers in the composition of the invention, wherein the number average molecular weight of the A-block is at least 1250 Da (as exemplified by the number average molecular weight of the A-block of 1500), gels are obtained at 37 C. that do not phase separate. Phase separation results in expulsion of free water/buffer from the interior of the gel-matrix, and shrinkage of the gel-matrix expressed as wet weight. Water-dissolved molecules will also exit the gel in large quantities with the expulsion of water-based solvents, hence leading to fast and high quantity release of the loaded water soluble molecules, which we refer to as burst or burst release. Avoiding early stage phase separation of the formulation is essential to keep water soluble drug-molecules inside the gel until they are released by diffusion and/or by gradual degradation of the gel-matrix.

(136) In conclusion the examples presented herein show that the block ratio of the PEG/PLCA tri-copolymer is important and the ratio needs to be greater than 1.2 but less than 2.8 in order to obtain a non-phase separating gel at 37 C. Furthermore, tri-block copolymers which are at least partially acylated have unexpected properties in a composition of the invention, when compared to compositions comprising unacylated tri-block copolymers. Notably the at least partial covalent modification of hydroxyl end groups of tri-block copolymers is very suitable for the sustained, controlled release of a pharmaceutically active ingredient, for example a water soluble drug, and results hardly in any burst release of water soluble molecules.

(137) Furthermore, a tri-block copolymer loaded with Hexabrix can be used for imaging in microCT studies but the contrast agent diffuses out of the tri-block copolymer within 1 day thereby preventing prolonged imaging of the tri-block copolymer. Surprisingly, a composition comprising a tri-block copolymer wherein at least part of the hydroxyl end-groups of the tri-block copolymer are covalently bound to a compound containing radiopaque atoms is suitable for microCT imaging over longer periods of up to time.

(138) It is shown herein that a composition comprising a tri-block copolymer wherein at least part of the hydroxyl end-groups of the tri-block copolymer are covalently bound to a compound containing radiopaque atoms maintains gel forming, gel erosion and controlled release properties whilst at the same time such composition can be visualized using micro CT.

(139) Furthermore it is shown that it is possible to visualize, e.g. using microCT or CT imaging, a blend of a tri-block copolymer wherein the hydroxyl end-groups of the tri-block copolymer are at least partially acylated with an acetyl group or wherein at least part of the hydroxyl end-groups of the tri-block copolymer are covalently bound to a compound containing radiopaque atoms, while maintaining gel forming, gel erosion and controlled release properties are maintained.