Therapeutic agents and delivery with microspheres

Abstract

The invention relates to methods for attaching therapeutic agents to structures produced by thermally induced phase separation as well as methods for coating devices and producing multi-layered microspheres.

Claims

1. A method for attaching a therapeutic agent to a hydrophobic microsphere structure produced by thermally induced phase separation comprising: i) at least partly submerging the hydrophobic microsphere structure in a culture medium; ii) subsequently contacting the hydrophobic microsphere structure with a wetting agent that wets the hydrophobic microsphere structure, wherein the wetting agent is ethanol which is added to the culture medium to a concentration of from 10% to 30%; and iii) attaching the therapeutic agent to the hydrophobic microsphere structure.

2. The method of claim 1 which results in the hydrophobic microsphere structure being fully submerged in the culture medium.

3. The method of claim 1 wherein the hydrophobic microsphere structure comprises poly(lactide-co-glycolide) (PLGA).

4. The method of claim 1 wherein the culture medium is Dulbecco's Modified Eagle's Medium (DMEM).

5. The method of claim 1 wherein the culture medium is tissue culture medium comprising serum.

6. The method of claim 1 wherein the hydrophobic microsphere structure is mixed with the culture medium before and after the contacting with the wetting agent.

7. The method of claim 1 further comprising incubating the hydrophobic microsphere structure at about 30 C. to about 40 C.

8. The method of claim 1 wherein the therapeutic agent is a cell.

9. The method of claim 1 wherein the therapeutic agent is an active pharmaceutical ingredient (API).

10. The method of claim 1 further comprising periods of static incubation interspersed with agitation.

11. The method of claim 10 comprising about 1 minute of agitation at about 300 rpm per hour of static incubation.

12. The method of claim 5 wherein the serum is fetal bovine serum (FBS) or fetal calf serum (FCS).

Description

(1) The invention will now be described in detail, by way of example only, with reference to the drawings, in which:

(2) FIG. 1 is a scanning electron microscope (SEM) image of a sectioned microsphere made in accordance with the invention from PLGA. The radial tubular pore structure can be seen.

(3) FIG. 2 is a higher magnification SEM image of the microsphere shown in FIG. 1.

(4) FIG. 3 is SEM images illustrating microspheres smaller than 10 m at (a) 600 and (b) 400 magnification.

(5) FIG. 4 illustrates microspheres incubated in (a) 500 l neat serum and (b) microspheres in 500 l culture medium containing 10% serum prior to addition of 300 l ethanol (1=50%, 2=60%, 3=70%, 4=80% v/v ethanol in deionised water).

(6) FIG. 5 illustrates that the wetting of TIPS microspheres by ethanol is dose dependent. PLGA microspheres were suspended in medium containing 20% FBS and treated with 0%, 35% (v/v), 70% (v/v) or 100% industrial methylated spirits (containing 99% ethanol [EtOH]), for 30 minutes, 3 hours and 24 hours under incubated culture conditions (37 C., 5% CO.sub.2, 95% humidity). Microsphere wetting was dependent on ethanol concentration and the duration of incubation (30 minutes, 3 hours, and 24 hours). Microspheres incubated in 70% (v/v) and 100% ethanol became immersed in medium (containing 20% FBS) after 3 hours, but those exposed to 35% (v/v) ethanol required 24 hours before they were completely submerged. These data demonstrate that PLGA microsphere wetting can be modified to various extents by altering ethanol concentrations and exposure times with this solvent.

(7) FIG. 6 illustrates that serum proteins adsorb to wetted TIPS microspheres in a dose-dependent manner.

(8) (A) PLGA microspheres were suspended in Dulbecco's Modified Eagle's Medium/F12 (1:1 v/v) containing 0%, 2%, 10% or 20% FBS and treated with 70% (v/v) ethanol (EtOH), for 30 minutes, 3 hours and 24 hours.

(9) (B) The total amount of serum protein adsorbed to microspheres was measured after each incubation period. Data points represent the mean (n=3S.E.M.) amount of total protein. *P0.05 indicate differences between wetting conditions.

(10) FIG. 7 illustrates that the adsorption of serum proteins to PLGA microspheres is dependent on the extent of pre-wetting.

(11) (A) PLGA microspheres were suspended in Dulbecco's Modified Eagle's Medium/F12 (1:1 v/v) containing 0%, 2% or 20% FBS and treated with 35% (v/v) or 70% (v/v) ethanol (EtOH), for 30 minutes, 3 hours and 24 hours. The total amount of serum protein adsorbed to microspheres was measured after each incubation period. Data points represent the mean (n=3S.E.M.) amount of total protein. *P0.05 indicate differences between wetting conditions.

(12) (B) Serum and EtOH treated microspheres were heated (95 C.) in Laemmli buffer containing the reducing agent -mercaptoethanol. Proteins were separated using 15% acrylamide gels and stained using silver nitrate. Bovine serum albumin (BSA) and fetal bovine serum (FBS) were included as external controls.

(13) FIG. 8 illustrates cell types shown to attach to the surface of PLGA TIPS microspheres. These include (a) endothelial progenitor cells, (b) smooth muscle cells, and (c) myoblasts.

(14) FIG. 9 shows human adipose derived mesenchymal stem cells attached to the surface of TIPS microspheres within 2 hours of incubation.

(15) FIG. 10 illustrates TIPS microspheres incubated with cells in LowBind plates attach most efficiently when incubated under static-dynamic conditions (1 hour static incubation followed by 1 min shaking at 300 rpm) (a). This results in the microspheres clustering towards the centre of flat-bottomed wells (b). Clustering of cellularised microspheres in the plates results in cells bridging between adjacent microspheres, forming clumps of microspheres (c-d). Inclusion of a curved low bind surface on the base of the plate prevents the microspheres clustering (e).

(16) FIG. 11 illustrates that (a) API (doxorubicin) does not bind to the surface of non-wetted, hydrophobic microspheres; (b) doxorubicin binds well to the surface of pre-wetted TIPS microspheres; (c) the unique TIPS microsphere structure of the wetted microspheres remains intact; and (d) controlled release of the API is observed over a prolonged period.

(17) FIG. 12 illustrates metallic surfaces prepared using the TIPS coating method.

(18) FIG. 13 shows a bare metal cobalt chromium coronary stent coated with TIPS microparticles.

(19) (a) Droplets of poly(lactide-co-glycolide) dissolved in dimethyl carbonate were deposited onto the stent via electrospraying under conditions that resulted in thermally induced phase separation. Shown at 210 magnification.

(20) (b) An individual TIPS microparticle present on the stent surface is shown at a higher magnification (6840).

(21) FIG. 14 illustrates a coaxial nozzle delivering two polymer solutions

(22) FIG. 15 illustrates a multi-layered microsphere bisected and imaged using scanning electron microscopy. The microsphere comprises an outer shell comprising 7.5% (w/v) PLGA in dimethyl carbonate and an inner core comprising 15% (w/v) PLGA in dimethyl carbonate.

(23) FIG. 16 illustrates SEMs showing TIPS microspheres produced using two different polymers delivered via a coaxial nozzle. Breakage of the distinct outer layer on the surface produced by polymer delivered from the outer nozzle reveals the different physical features of the inner layer produced by polymer delivered from the inner nozzle. The added line indicates the boundaries between the two layers. (a) Inner nozzle: 4% PEG (polyethylene glycol) dissolved in DMC (dimethyl carbonate) and filtered with 0.45 m cellulose syringe filters Outer nozzle: 5% PLGA polymer (75:25) PURASORB PDLG 7502 dissolved in DMC (b-c) Inner nozzle: 4% PLGA polymer (75:25) PURASORB PDLG 7502 dissolved in DMC Outer nozzle: 5% PEG dissolved in DMC and filtered with 0.45 m cellulose syringe filters (d) Inner nozzle: 4% PEG dissolved in DMC and filtered with 0.45 m cellulose syringe filters Outer nozzle: 5% PLGA polymer (75:25) PURASORB PDLG 7502 dissolved in DMC (e) Inner nozzle: 4% PLGA polymer (75:25) PURASORB PDLG 7502 dissolved in DMC Outer nozzle: 5% PEG dissolved in DMC and filtered with 0.45 m cellulose syringe filters (f-g) Inner nozzle: 5% PEG dissolved in DMC and filtered with 0.45 m cellulose syringe filters Outer nozzle: 4% PLGA polymer (75:25) PURASORB PDLG 7502 dissolved in DMC

EXAMPLES

Example 1

Production of Microspheres Using the Thermally Induced Phase Separation Process (TIPS Microspheres)

(24) Poly(D,L-lactide-co-glycolide) (PLGA) (75:25) (Medisorb, Alkermes, USA) was used as the polymeric matrix, dissolved in dimethyl carbonate (of >99.9% purity, Sigma Aldrich, UK). PLGA was dissolved in dimethyl carbonate (DMC) at 1:6 w/v (0.833 g PLGA was dissolved in 5 ml DMC for 2 h in a 25 ml Falcon tube, under magnetic stirring). The polymer solution was dripped from a syringe fitted with various sized needle orifices, into liquid nitrogen to rapidly induce the phase separation. Each drop of polymer solution was allowed to equilibrate to the liquid nitrogen temperature, demarked by sinking, prior to the addition of further drops to prevent microsphere agglomeration during processing. The frozen spheres were subsequently freeze-dried overnight to yield the TIPS microspheres. TIPS microspheres were sectioned using a Wilkinson Sword Razor blade to permit examination of the interior pore structure by scanning electron microscopy (SEM).

(25) The pore structure is highly interconnected with a structure typical of such TIPS foams. Specifically the DMC solvent has a freezing temperature of 1 C. and if the polymer solution is frozen rapidly using liquid nitrogen, tubular pores develop due to the crystallisation front of the freezing solvent. Significantly here, the freeze front is from the outside in; therefore a radial pore structure of tubular pores, interconnected by a ladder-like structure of smaller pores occurred, as shown in FIGS. 1 and 2.

(26) The size of the microspheres is related to the size of the needle orifice, smaller needle orifices give smaller microspheres, as shown in the Table 1, below.

(27) TABLE-US-00001 TABLE 1 Effect of needle orifice size on microsphere size Needle orifice size ~Microsphere size 700 m 1.7 mm 350 m 1.2 mm 200 m 900 m

(28) FIG. 3 illustrates microspheres which are smaller than 10 m, created with a needle orifice size of approximately 690 m using an electrospraying process.

(29) TIPS microsphere fabrication using dimethyl carbonate as a solvent and rapid quenching in liquid nitrogen resulted in highly ordered interconnected porosity, with radial pores (channel-like) produced from the advancement of the solvent crystallisation front towards the centre of the sphere (parallel to the direction of heat transfer) for a neat PLGA TIPS microsphere. During TIPS the solution is separated into a polymer-rich phase and a polymer-lean phase due to the crystallisation of the solvent, when the temperature of the polymer solution is lower than the freezing point of the solvent and the polymer is expelled from the crystallisation front to form a continuous polymer-rich phase. The solvent is sublimed to leave the pores, which are a three-dimensional fingerprint of the geometry of the solvent crystals. At higher magnification the structure of the neat PLGA TIPS microsphere is observed to have a highly anisotropic channel-like morphology with an internal ladder-like structure, which is a characteristic morphology of foams formed by solid-liquid TIPS. The exterior of the neat PLGA microspheres, composite and protein encapsulated TIPS microspheres consist of a skin region of about 2 m thickness with a smooth polymer surface, peppered with pores of 1 to 5 m and covered with chevron like patterns due to the initial freeze front of the solvent across the droplet surface. Once the freeze fronts progress towards the centre of the droplet, the pore structure becomes more ordered, interconnected and ladder-like. The size of the spheres can be controlled by the size of the needle orifice, with smaller spheres produced from needles of narrower orifice (Table 1). The microspheres are monodisperse due to the consistent droplet formation. Voids are evident in the samples and are due to the entrapment of air during the manual droplet formation method, and the short drop distance to the liquid nitrogen used in the current study. The voids consist of a neck extending from the exterior surface of the sphere. Formation of these air pockets might be prevented by the use of a vibrating needle and a more optimized processing technique. The microstructure of the pores and walls can be controlled by varying the polymer concentration, filler loading content, quenching temperature and solvent used. Porosity increases with decreasing polymer concentration and filler content. Foams of up to 95% porosity can be achieved using the TIPS technique. Our method using DMC as solvent and PLGA polymer enables the formation of a slightly dense skin region and radial pores, thereby enhancing the mechanical properties over a random pore structure.

(30) The use of an electromagnetic vibrating needle may be employed to i) maintain dispersion of the particulates in the polymer solution, ii) prevent blocking of the needle and iii) achieve smaller microspheres (100 to 800 m) by vibrating the nozzle itself. The deviations in sphere size will depend on the density and surface tension of the matrix. Roughly, the smallest achievable drop diameter is 1.5 to 2 times larger than the nozzle diameter used.

Example 2

Wetting of TIPS Microspheres

(31) TIPS microspheres were wetted using the following protocol: 1. Approximately 30 mg of dry TIPS microspheres were placed into a 1.5 ml microfuge tube. 2. 500 l of foetal calf serum (or tissue culture medium (e.g. Minimum Essential Medium Eagle [Sigma Aldrich M2279]+10% foetal calf serum) was added to microfuge tube. 3. The microfuge was vortexed for 20 seconds to mix the microspheres with the solution. 4. 300 l of ethanol (diluted to 50, 60, 70 and 80% v/v with de-ionized water) was added to each tube. 5. The microfuge was vortexed for 20 seconds to mix the microspheres with the solution. 6. The microfuge tubes were placed in an incubator at 37 C. for 90 minutes and 240 minutes. 7. The incubation solution was removed from the microfuge tubes and replaced with fresh culture medium appropriate for the cells to be attached.

(32) FIG. 4 illustrates microspheres incubated in (a) 500 l neat serum and (b) microspheres in 500 l culture medium containing 10% serum prior to addition of 300 l ethanol (1=50%, 2=60%, 3=70%, 4=80% v/v ethanol in deionised water). The majority of microspheres sunk within 90 minutes incubation at 37 C. The effect is more pronounced at concentrations of ethanol above 60% v/v. All of the microspheres sunk within 4 hours of incubation.

(33) FIG. 5 illustrates that the wetting of TIPS microspheres by ethanol is dose dependent. PLGA microspheres were suspended in medium containing 20% FBS and treated with 0%, 35% (v/v), 70% (v/v) or 100% industrial methylated spirits (containing 99% ethanol [EtOH]), for 30 minutes, 3 hours and 24 hours under incubated culture conditions (37 C., 5% CO.sub.2, 95% humidity). Microsphere wetting was dependent on ethanol concentration and the duration of incubation (30 minutes, 3 hours, and 24 hours). Microspheres incubated in 70% (v/v) and 100% ethanol became immersed in medium (containing 20% FBS) after 3 hours, but those exposed to 35% (v/v) ethanol required 24 hours before they were completely submerged. These data demonstrate that PLGA microsphere wetting can be modified to various extents by altering ethanol concentrations and exposure times with this solvent.

(34) FIG. 6 illustrates that serum proteins adsorb to wetted TIPS microspheres in a dose-dependent manner.

(35) (A) PLGA microspheres were suspended in Dulbecco's Modified Eagle's Medium/F12 (1:1 v/v) containing 0%, 2%, 10% or 20% FBS and treated with 70% (v/v) ethanol (EtOH), for 30 minutes, 3 hours and 24 hours.

(36) (B) The total amount of serum protein adsorbed to microspheres was measured using the micro Lowry protein assay after each incubation period. Data points represent the mean (n=3S.E.M.) amount of total protein.

(37) *P0.05 indicate differences between wetting conditions.

(38) FIG. 7 illustrates that the adsorption of serum proteins to PLGA microspheres is dependent on the extent of pre-wetting.

(39) (A) PLGA microspheres were suspended in Dulbecco's Modified Eagle's Medium/F12 (1:1 v/v) containing 0%, 2% or 20% FBS and treated with 35% (v/v) or 70% (v/v) ethanol (EtOH), for 30 minutes, 3 hours and 24 hours. The total amount of serum protein adsorbed to microspheres was measured using the micro Lowry protein assay after each incubation period. Data points represent the mean (n=3S.E.M.) amount of total protein. *P0.05 indicate differences between wetting conditions.

(40) (B) Serum and EtOH treated microspheres were heated (95 C.) in Laemmli buffer containing the reducing agent -mercaptoethanol. Proteins were separated using 15% acrylamide gels and stained using silver nitrate. Bovine serum albumin (BSA) and fetal bovine serum (FBS) were included as external controls.

Example 3

Cell Attachment to TIPS Microspheres

(41) Cells were attached to TIPS microspheres which were wetted using the method described in Example 1 using the following protocol: 1. Wetted TIPS microspheres in fresh culture medium were transferred to wells of a Corning Costar Ultra-Low attachment multiwell plate. 2. The culture medium was removed leaving the only wetted microspheres in the wells of the plate. 3. 400 l of fresh tissue culture medium was added to the wells. 4. 100 l of medium containing 110.sup.5 cells was added to the wells. 5. The plate was placed on an orbital shaker (STUART SSM5) placed inside a CO.sub.2 incubator at 37 C. The shaker was set to agitate the plate at 300 rpm for 1 minute per hour. 6. The plate was incubated for 18 hours before the cellularised microspheres are removed. 7. Cells were stained and imaged by fluorescence microscopy as shown in FIG. 4.

(42) FIG. 8 illustrates cell types shown to attach to the surface of PLGA TIPS microspheres. These include (a) Muller stem cells, (b) endothelial progenitor cells, (c) smooth muscle cells, (d) lung epithelial cells, (e) mesangioblasts, (f) myoblasts.

(43) Cell attachment and delivery both dependent on porosity. Each cell type is suited to different surface textures for attachment, spreading and differentiation. Table 2 illustrates the preferred size ranges and porosity for various clinical conditions.

(44) TABLE-US-00002 TABLE 2 Preferred microsphere size ranges and porosity for various clinical conditions Size Range Clinical Condition (m) Porosity Cell Type Delivery Route Fistula 100-400 Medium Mesenchymal stem Into fistula tract cells (MSC; via syringe and autologous or cannula allogenic) Vision degeneration <100 Medium Mller stem cells Injection, spray Large orthopaedic defects 10-300 Low Osteoblasts, Packing, Medium chondrocytes and spraying, syringe mesenchymal stem and cannula cells Tumour immobilisation and <10 Medium Dendritic cells, T, Intravascular immunotherapy High lymphocytes Osteoarthritis 10-300 Low MSC Injection, spray Medium (intra-articular) and intravenously Incontinence 100-400 Medium Myoblasts, MSC Injection Heart failure <100 Medium Bone marrow, Intravenously, High cardiopoietic cells, catheter, cardiac progenitor intramyocardial cells Muscular dystrophy <100 Low MSC, Injection, Medium mesangioblast, intravenously, myoblast packing 3D models and cell 10-400 Low Adherent cells Expansion or in expansion Medium vitro modelling High in suspension cultures Fermentation/vaccine 10-500 Low Adherent cells, Alternative to production/brewing/filtration Medium microorganisms wave bag/cell High bag bioreactors, biofiltration medium.

(45) The use of a low attachment plate during the cell attachment protocol outlined above prevents non-adhered cells attaching to the plate instead of the microspheres. The period of agitation re-suspends cells to increase the likelihood of cells attaching to the sunk microspheres. Experiments show cells attach to TIPS microspheres within 2 hours incubation using the wetting and agitation process. FIG. 9 shows human adipose derived mesenchymal stem cells attached to the surface of TIPS microspheres within 2 hours of incubation.

Example 4

Inclusion of a Curved Low Bind Surface on the Base of the Tissue Culture Plate

(46) Use of a tissue culture plate comprising a barrier of non-adherent material ascending from the base of the wells can reduce clumping of cellularised TIPS microspheres towards the centre of wells in the culture plate. Clumping can be caused by vortexing of the culture medium in the dynamic phase of incubation.

(47) In one embodiment, the tissue culture plate comprises an upward curve of non-adherent silicone material attached to the centre of the well of the low bind tissue culture plate described in Example 3. This prevents the microspheres from clustering and adhering together as illustrated in FIG. 10.

(48) FIG. 10 illustrates TIPS microspheres incubated with cells in LowBind plates attach most efficiently when incubated under static-dynamic conditions (1 hour static incubation followed by 1 min shaking at 300 rpm) (a). This results in the microspheres clustering towards the centre of flat-bottomed wells (b). Clustering of cellularised microspheres in the plates results in cells bridging between adjacent microspheres, forming clumps of microspheres (c-d). Inclusion of a curved low bind surface on the base of the plate prevents the microspheres clustering (e).

Example 5

Attachment of Active Pharmaceutical Ingredients (API) to TIPS Microspheres

(49) The wetting technique described in Example 2 can facilitate the attachment of (API), which are soluble in aqueous solution to TIPS microspheres.

(50) 100 mg of TIPS microspheres was added to a 7 ml plastic container. The microspheres were wetted with 70% ethanol for 1 minute. The alcohol was removed and the microspheres washed with 5 ml deionised water. 1 ml of doxorubicin (0.5 mg/ml in water) was added to the container. The microspheres and doxorubicin solution were mixed by pipetting and incubated overnight at room temperature with agitation at 150 rpm on an orbital shaker. Control non-wetted microspheres were treated in the same manner. The microspheres were centrifuged at 13000 rpm for 20 seconds. The doxorubicin remained bound to the surface of the wetted microspheres whereas it was separated from the non-wetted samples.

(51) FIG. 11 illustrates that (a) API (doxorubicin) does not bind to the surface of non-wetted, hydrophobic microspheres; (b) doxorubicin binds well to the surface of pre-wetted TIPS microspheres; (c) the unique TIPS microsphere structure of the wetted microspheres remains intact; and (d) controlled release of the API is observed over a prolonged period.

Example 6

Coating of a Metallic Scaffold

(52) The following example describes the method for coating a metallic scaffold (hereafter termed scaffold): 1. The scaffold material was immersed into a solution of 5 wt % poly(lactide-co-glycolide) dissolved in dimethyl carbonate for 1 minute. 2. The coated scaffold material was immediately placed into a container of liquid nitrogen resulting in separation of the polymer solution into polymer rich and polymer lean phases. 3. The frozen coated scaffold was placed into a freeze drier and lyophilized for 24 hours until removal of the solvent was complete. 4. The coated scaffold was imaged using scanning electron microscopy (FIG. 12).

(53) FIG. 13 shows a bare metal cobalt chromium coronary stent coated with TIPS microparticles. Droplets of poly(lactide-co-glycolide) dissolved in dimethyl carbonate were deposited onto the stent via electrospraying under conditions that resulted in thermally induced phase separation. After droplets were deposited by electrospraying, the method of FIG. 12 from step 2 onwards was followed.

Example 7

Creation of Multi-Layered TIPS Microspheres

(54) The following method is used to create multi-layered microspheres consisting of an outer shell comprising 7.5% (w/v) poly(lactide-co-glycolide) (PLGA) in dimethyl carbonate and an inner core comprising 15% (w/v) PLGA in dimethyl carbonate (DMC). 1. Solutions of PLGA (7.5 wt % or 15 wt % in DMC) were loaded into separate 10 ml syringes. 2. The two syringes containing the polymer solutions were attached to a syringe pump and delivered via tubing at a rate of 3 ml/min to a Nisco Var D Classic open electromagnetically driven single nozzle encapsulator unit fitted with a stainless steel coaxial nozzle consisting of two subnozzles positioned coaxially (FIG. 14). The encapsulator unit was set to 1.80 kHz and 100% amplitude. 3. The polymer droplets were collected in a liquid nitrogen quenching bath, transferred to a container and lyophilized for 24 hours until removal of the solvent was complete. 4. The multi-layered microspheres were bisected and imaged using scanning electron microscopy (FIG. 15).

(55) FIG. 16 illustrates further SEM images of microspheres produced using the method described above.

(56) All cited references are herein incorporated in their entirety.