CA2223583A1 - Device for releasing aggregation-stabilized, biologically active agent - Google Patents

Abstract

A device for the sustained release in vivo of a water soluble, biologically active agent wherein the agent is susceptible to aggregation comprising a drug delivery device and aggregation-stabilized, biologically active agent wherein the aggregation-stabilized agent is disposed within the drug delivery device.

Description

CA 02223~83 1997-12-04 DEVICE FOR RELEASING
AGGREGATION-STABILIZED, BIOLOGICALLY ACTIVE AGENT

~ Related A~plications This application is a Continuation-in-Part of co-pending PCT Patent Application PCT/US95/07348 filed June 7, 1995, which designates the United States (and is a Continuation-in-Part application cl~;m~ng priority to U.S.
Application Serial No. 08/279,784 filed July 25, 1994);
U.S. Patent Application Serial No. 08/478,502, filed June 7, 1995; U.S. Patent Application Serial No. 08/483,318, filed June 7, 1995; U.S. Patent Application Serial No. 08/473,544 filed June 7, 1995; and U.S. Patent Application Serial No. 08/477,725, filed June 7, 1995, which are incorporated herein by reference in their entirety.

Backqround of the Invention Many illnesses or conditions require ~m; n;stration of a constant or sustained level of a medicament or biologically active agent to provide the most e~fective prophylactic or therapeutic. This may be accomplished through a multiple dosing regimen or by employing a system that releases the medicament in a sustained fashion.
Attempts to sustain medication levels include the use of biodegradable materials, such as polymeric matrices, cont~;n;ng the medicament. The use of these matrices, for example, in the form of microparticles or microcarriers, provides an improvement in the sustained release of medicaments by utilizing the inherent biodegradability of the polymer to control the release of the medicament and provide a more consistent, sustained level of medication and improved patient compliance.

CA 02223~83 1997-12-04 WO 96/40074 PCT/U',GI'~ 6 However, these sustained release devices often exhibited high initial bursts of agent release and m;n;m~71 agent release thereafter. Further, due to the high solution concentration of agent within and localized around these sustained release devices, the agent molecules have tended to aggregate thereby increasing ;mm77nogenicity in vivo and interfering with the desired release profile for the agent.
Therefore, a need exists for a means for sustaining lC the release of a biologically active agent in vivo without significant aggregate formation and thus with a reduced immune response to the agent over the release period of the agent.

Summary of the Invention This invention relates to a device for the sustained release in vivo of a water soluble, biologically active agent wherein said agent is susceptible to aggregation, comprising a drug delivery device and aggregation-stabilized, biologically active agent wherein the aggregation-stabilized agent is disposed within the drug delivery device.
There are many advantages to this sustained release device for a biologically active agent. These advantages include longer, more consistent in vivo blood levels of the agent, lower initial bursts of the agent, and increased therapeutic benefits by eliminating fluctuations in serum agent levels. The advantages also include better retention of biological activity of the agent and reduced immunogenicity when in vivo. The advantages further include more complete release of an agent from a sustained release device.

CA 02223~83 1997-12-04 W O 96/40074 PCTrUS96/08526 Brief Descri~tion of the Drawinqs Figure 1 is a plot of a) the cumulative release of monomeric erythropoietin (EPO), b) the cumulative release of EPO (monomer EPO plus aggregated EPO), and c) the percentage of EPO which is released as a monomer during the interval between an indicated time point and the immediately preceding time point, in vitro in HEPES buffer, from microcarriers of unblocked poly(lactide-co-glycolide) polymer (PLGA) ~10,000 Dalton MW), containing 10~ (w/w) MgCO3 and 5~ (w/w) of the Aml formulation of Example 6, versus time over a 28 day interval.
Figure 2 is a plot of a) the cumulative release of monomeric EPO, b) the cumulative release of EPO (monomer plus aggregate), and c) the percentage of EPO which is released as a monomerduring the interval between an indicated time point and the immediately preceding time point, in vitro in HEPES buffer, from microcarriers unblocked PLGA (10,000 Dalton MW), containing 10~ (w/w) MgCO3 and 5~ (w/w) of the Am7 formulation of Example 6, versus time over a 28 day interval.
Figure 3 is a plot of a) the cumulative release of monomeric EPO, b) the cumulative release of EPo (monomer plus aggregate), and c) the percentage of EPO which is released as a monomer during the interval between an indicated time point and the immediately preceding time point, in vitro in HEPES buffer, from microcarriers of blocked PLGA (10,000 Dalton MW), containing 10~ (w/w) ZnC03 and 10~ (w/w) of the Znl formulation of Example 6, versus time over a 28 day interval.
Figure 4 is a plot of the serum concentration (IU/ml) of Interferon-~,2b (IFN-~,2b) in rats, which were subcutaneously administered IFN-~,2b controlled release formulated microcarriers of Example 2, versus time over a 6 day interval.

CA 02223~83 1997-12-04 WO 96/40074 PCTAUSg5.'C-'~

Figure 5 is a plot of the serum concentration (IU/ml) of IFN-~,2b in rats, which were subcutaneously administered IFN-~,2b controlled release Formula 2 microcarriers of Example 2, versus time over a 6 day interval.
Figure 6 is a plot of the serum concentration (IU/ml) of IFN-~,2b in rats, which were subcutaneously administered IFN-~,2b controlled release Formula 3 microcarriers of Example 2, versus time over a 7 day interval.
Figure 7 is a plot of the serum concentration (IU/ml) of IFN-a,2b in rats, which were subcutaneously administered IFN-~,2b controlled release Formula 4 microcarriers of Example 2, versus time over a 7 day interval.
Figure 8 is a plot of the serum concentration (IU/ml) of IFN-~,2b, in rats, which were subcutaneously administered IFN-~,2b controlled release Formula 5 microcarriers of Example 2, versus time over a 7 day interval.
Figure 9 i8 a plot of the serum concentration (IU/ml) of IFN-~,2b in rats, which were subcutaneously administered IFN-~,2b controlled release Formula 6 microcarriers of Example 2, versus time over a 7 day interval.
Figure 10 is a plot of the serum concentration (IU/ml) of IFN-~,2b versus time over a 7 day interval in rats which were subcutaneously ~m; n; stered IFN-~,2b controlled release Formula 7 microcarriers of Example 2 having a 1:1 zinc carbonate-to-IFN-~,2b ratio.
Figure 11 is a plot of the serum concentration (IU/ml) of IFN-~,2b versus time over a 29 day interval in rats which were subcutaneously ~m; n; stered a) IFN-~,2b controlled release microcarriers of Formula 8 of Example 2, wherein the rats were ;mmllnosuppressed with cyclosporin A
and hydrocortisone (two groups) and b) the same formulation of IFN-~,2b controlled release microcarriers wherein the rats were not immunosuppressed.

CA 02223~83 1997-12-04 Figure 12 is a plot of the serum concentrations (IU/ml) of IFN-~,2b versus time over a 14 day interval in monkeys which were subcutaneously administered a) IFN-~,2b controlled release microcarriers of Example 2 having a 1:8 zinc carbonate to IFN-~,2b ratio and b) an equal dose of IFN-~,2b in 0.9% saline solution.
Figure 13 is a plot of the serum concentration (ng/ml) of hGH versus time over a 28 day interval in rats which were subcutaneously administered a) aggregation-stabilized hGH microcarriers of 31K unblocked PLGA containing 1~ ZnC03 of Example 5 wherein the rats were immunosuppre~csed with cyclosporin A and hydrocortisone and b) the same hGH
microcarriers wherein the rats were not immunosuppressed.
Figure 14 is a plot of the serum concentration (ng/ml) of hGH versus time over a 28 day interval in rats which were subcutaneously administered a) aggregation-stabilized hGH microcarriers of 8K unblocked PLGA containing 1~ ZnC03 of Example 5 wherein the rats were immunosuppressed with cyclosporin A and hydrocortisone and b) the same hGH
microcarriers wherein the rats were not immunosuppressed.
Figure 15 is a plot of the serum concentration (ng/ml) of hGH versus time for a 61 day interval in monkeys which were subcutaneously administered aggregation-stabilized hGH
microcarriers of Example 5 cont~;n;ng 15~ hGH (complexed with zn+2 at a 6:1 Zn+2:hGH molar-ratio), 6~ w/w ZnC03 and lOK blocked PLGA.
Figure 16 is a plot of the serum concentration (ng/ml) of hGH versus time for a 60 day interval in monkeys which were subcutaneously administered aggregation-stabilized hGH
microcarriers of Example 5 containing 15~ hGH (complexed with Zn+2 at a 6:1 Zn+2:hGH molar ratio), 1~ w/w ZnC03 and 8K unblocked PLGA.
Figure 17 is a plot of the serum concentration (ng/ml) of hGH versus time for a 68 day interval in monkeys which were subcutaneously administered aggregation-stabilized hGH

CA 02223~83 1997-12-04 WO 96/40074 PCTAU~96'Q~'~6 microcarriers of Example 5 containing 15~ hGH tcomplexed with zn+2 at a 6:1 Zn+2:hGH molar ratio), 1~ w/w ZnCO3 and 3lK unblocked PLGA.
Figure 18 is a plot of the serum concentration (ng/ml) of hGH and IGF-l versus time for a 32 day interval in monkeys which were subcutaneously administered aggregation-stabilized hGH microcarriers of Example 16 in 8K unblocked PLGA.
Figure 19 is a plot of the serum concentration (ng/ml) of hGH versus time for 30 and 39 day intervals for a) aggregation-stabilized hGH 8K unblocked PLGA
microcarriers and b) daily aqueous hGH injections, respectively.
Figure 20 is a plot of the percent reticulocytes in blood of cyclosporin/hydrocortisone (CS/HC) treated and untreated rats, which were subcutaneously injected with 10,000 units of the EPO sustained release microcarriers RMAm7, described in Example 17 a bolus of 2,000 units of aqueous EPO, administered on day 28, respectively, versus time over a 36 day interval.
Figure 21 is a plot of the serum concentration (IU/ml) of EPO in rats, which were subcutaneously administered various EP0 sustained release microcarriers, described in Example 6, versus time over a 22 day interval.
Figure 22 is a plot of the percent reticulocytes in blood of rats, which were subcutaneously injected with 10,000 units of various EPO sustained release microcarriers, described in Example 6, versus time over a 28 day interval.
Figure 23 is a plot of the serum concentration (IU/ml) of IFN-~,2b versus time over a 7 day interval in rats which were subcutaneously administered three different IFN-~,2b controlled release microcarriers of Example 2 having zinc carbonate to IFN-~,2b ratios of 1:1, 3:1 and 8:1.

CA 02223~83 1997-12-04 WO 96/40074 PCTrUS96/08526 Detailed Descri~tion of the Invention A biologically active agent, as defined herein, is an agent, or its pharmaceutically acceptable salt, which is in its molecular, biologically active form when released in vivo, thereby possessing the desired therapeutic and/or prophylactic properties in vivo. Biologically active agents suitable for the composition and method of the invention are agents which are soluble in aqueous solutions and biological fluids and which are susceptible to aggregation in vivo. Examples of suitable biologically active agents include proteins such as ;mmllnoglobulin-like proteins, antibodies, cytokines (e.g., lymphokines, monokines, chemokines), interleukins, interferons, erythropoietin, nucleases, tumor necrosis factor, colony stimulating factors, insulin, enzymes, tumor suppressors, hormones (e.g., growth hormone and adrenocorticotropic hormone), antigens (e.g., bacterial and viral antigens) and growth factors; peptides such as protein inhibitors;
nucleic acids, such as antisense molecules;
oligonucleotides; and ribozymes.
A sustained release of a biologically active agent is a release which results in biologically effective serum levels of the biologically active, molecular (monomeric or non-aggregated) form of the agent over a period longer than that obtained following direct administration of an aqueous solution of the agent. A biologically effective serum level of an agent is a level which will result in the desired biological response within the recipient. Usually, in a sustained release, the serum level of the agent is above endogenous levels. Typically, a sustained release of an agent is for a period of a week or more, and preferably for two weeks or more.
A sustained release of non-aggregated, biologically active agent can be a continuous or non-continuous release CA 02223~83 1997-12-04 with relatively constant or varying rates of release from a drug delivery device. The continuity of relea~e of the biologically active agent can be affected by the loading of the agent, selection of excipients to produce the desired effect, and/or by other conditions such as the type of polymer used if the biologically active agent is encapsulated within a polymeric matrix.
A drug delivery device, as defined herein, includes any composition, such as diffusion-controlled polymeric and protein systems of the reservoir or matrix-type, or systems such as pressure-driven osmotic or syringe pumps wherein the rate of release of a biologically active agent is sustained by use of a drug delivery device to release said agent in vivo.
Aggregation-stabilized biologically active agent, as defined herein comprises a suitable agent in its biologically active, molecular (monomeric) form wherein the biologically active agent is stabilized against aggregation during formation of the sustained release device and while the device is employed in vivo. A biologically active agent can be aggregation-stabilized by several means, such as by controlling the solubilization of the agent in vivo and by controlling the environmental conditions experienced by the agent during device formation and in vivo. These means are typically dependent upon the specific biologically active agent to be aggregation-stabilized.
Preferably, the means for aggregation-stabilizing a biologically active agent should not convert the agent to a form that will reduce in vivo biological activity such as by oxidation.
An aggregation-stabilized biologically active agent is stabilized against significant aggregation in vivo over the sustained release period. Significant aggregation is defined as an amount of aggregation that will reduce or preclude the achievement of effective serum levels in vivo CA 02223~83 1997-12-04 WO 96/40074 PCT/U',.'~3~6 of the biologically active agent over the sustained release period. Typically, significant aggregation is aggregation of about 10~ or more of the original amount of biologically active agent in the sustained drug delivery device.
Pre~erably, aggregation is maintained below about 5~ o~ the initial loading of the molecular form of the agent. More preferably, aggregation is maintained below about 2~ of the initial loading of biologically active agent.
In one embodiment of the sustained release device of the present invention, the biologically active agent is mixed with an aggregation-stabilizer wherein the in vivo solubilization of the biologically active agent is controlled. Typically an aggregation-stabilizer reduces the solubility of the biologically active agent, precipitates out a salt of the agent or forms a complex o~
the agent. The aggregation-stabilizer and the biologically active agent can be separately contained within the sustained drug delivery device, such as a device containing particles of aggregation-stabilizer and separate particles o~ biologically active agent, and/or can be combined together in complexes or particles which contain both the aggregation-stabilizer and the biologically active agent.
The suitability o~ candidate aggregation-stabilizers for stabilizing a biologically active agent against aggregation can be determined by one o~ ordinary skill in the art by performing a variety of stability indicating techniques such as SEC, polyacrylamide gel electrophoresis (PAGE) and potency tests on protein obtained from particles containing the aggregation-stabilized agent and for the duration of release from the sustained release device, as described in Example 5 ~or hGH and Examples 8-9 for EPO.
Suitable particles o~ aggregation-stabilized biologically active agent are solid particles, including lyophilized particles, freeze-dried particles, pressed pellets, and particles formed by any other means known in CA 02223~83 1997-12-04 WO 96/40074 PCT/U',61'~~~?6 the art for forming a solid particle from a mixture of two components (e.g., biologically active agent and an aggregation stabilizer) wherein one component is temperature sensitive.
The amount of an agent which is contained in a sustained release device containing biologically active, aggregation-stabilized particles of the agent is a therapeutically or prophylactically e~fective amount which can be determined by a person of ordinary skill in the art taking into consideration factors such as body weight, condition to be treated, type of device used, and release rate from the device.
In one example of this embodiment wherein the in vivo solubilization of a biologically active agent is controlled, a biologically active agent is aggregation-stabilized when mixed with at least one type o~ metal cation from a metal cation component, which is the aggregation-stabilizer, wherein the agent is complexed and/or complexes in vivo with the metal cation to aggregation-stabilize the agent.
Suitable aggregation-stabilizing metal cations include biocompatible metal cations which will not significantly oxidize the agent. Typically, oxidation of a biologically active agent by a metal cation is not significant if this oxidation results in a loss of the agent's potency of about 10~ or less. A metal cation component is biocompatible if it is non-toxic to the recipient in the quantities used, and also presents no significant deleterious or untoward effects on the recipient's body, such as an immunological reaction at the injection site. Preferably, the metal cation is multivalent.
Examples of suitable aggregation-stabilizing metal cations include cations of non-transition metals, such as Mg+2 and Ca+2. Suitable aggregation-stabilizing metal cations also include cations of transition metals, such as CA 02223~83 1997-12-04 W 096/40074 PCT~US96/08526 Cu+2, Co+2, Fe+3 and Ni+2. In a preferred embodiment, zn+2 is used as an aggregation-stabilizing metal cation. The suitability of metal cations for stabilizing a biologically active agent can be determined by one of ordinary skill in the art by performing a variety of stability indicating techniques such as polyacrylamide gel electrophoresis, isoelectric focusing, reverse phase chromatography, size exclusion chromatography (SEC) and potency tests on particles of the biologically active agent containing metal cations to determine the potency of the agent after par~icle formation, such as by lyophilization, and for the duration of release from microparticles.
It is preferred that the metal cation and biologically active agent are complexed within the sustained drug delivery device before administration to a subject.
It is also preferred that the mixture of the metal cation and the biologically active agent are in the form of solid particles, more preferably, lyophilized particles.
The molar ratio of metal cation to biologically active agent is typically between about 1:2 and about 100:1, and is preferentially between about 2:1 and about 10:1.
The use of metal cations to form aggregation-stabilized particles of the biologically active agents, interferon (IFN) and human growth hormone (hGH), are 2' further described in Examples 1 and 4. In addition, the formation o~ sustained release devices o~ polymeric microcarriers cont~;n;ng metal cation-stabilized IFN or hGH
are described in Examples 2 and 5. Furthermore, the aggregation-stabilization ef~icacy o~ metal cations complexed with IFN or hGH, within lyophilized particles dispersed in polymeric microcarriers, over a sustained release period in vivo are described in Examples 10-12 or Examples 13-16, respectively.
The use of additional metal cations, dispersed within the polymeric matrix of a sustained release device, to CA 02223~83 1997-12-04 WO 96/40074 PCTAUS96/~ 6 further aggregation-stabilize a biologically active agent (hGH or IFN) are described in Examples 14 and 18.
The polymeric matrix is believed to function as a reservoir of metal cations so that the formation of cation-complexed protein is favored and dissociation into solubleprotein is disfavored. ~herein the aqueous solubility of the metal cation component in the polymeric matrix is low, the release of metal cations from the matrix is slow, thus modulating the solubility of the protein.
In another example of the embodiment wherein the solubility of a biologically active agent is reduced by an aggregation stabilizer, the biologically active agent is mixed with an aggregation stabilizer which reduces solubility by precipitating the agent from the aqueous solution, thereby maint~;n;ng a suitably low localized concentration of soluble agent below a concentration at which significant aggregation occurs. A localized concentration of an agent is the concentration of solvated agent within, between or immediately surrounding the sustained release device. Suitable materials for precipitating an agent, such as a protein, without denaturing the agent, include salts which are in the Hofmeister series of precipitants of serum globulins (or "salting-out salts") as described by Thomas E. Creighton in Proteins: Structures and Molecular Principles, pl49-150 (published by W.H. Freeman and Company, New York).
Suitable salting-out salts for use in this invention include, for example, salts cont~;n;ng one or more of the cations Mg+2, Li+, Na+, K+ and NH4+; and also contain one or more of the anions so4~2, HPo4-2~ acetate, citrate, tartrate, Cl-, N03-, Cl03-, I-, Cl04- and SCN-.
Again, the ~iologically active agent and the precipitant can be combined within particles and/or can be separately contained within the sustained release device.
Preferably, a biologically active agent and a precipitant CA 02223~83 1997-12-04 WO 96/40074 PCTrUS96/08526 are combined in a lyophilized particle. The formation of lyophilized particles cont~;n;ng the agent erythropoietin and a precipitant, and the use of these particles in polymeric microcarrier sustained release devices, are described in Examples 6 and 7. The efficacy of precipitants in preventing aggregation of EPO in vitro and in vivo over a sustained period are also described in Examples 8-9 and Example 17, respectively.
In yet another embodiment for stabilizing a biologically active agent against aggregation, the agent is mixed with a buffer which will maintain the agent under pH
conditions in vivo that can affect the rate of solubilization of the agent and/or prevent the formation in vivo of biologically inactive or insoluble forms (precipitates or gels which are insoluble in vivo) of the agent. Examples of such buffers include, for instance, phosphate buffers.
A pre~erred sustained release device of the present invention is a biocompatible polymeric matrix containing particles o~ an aggregation-stabilized biologically active agent dispersed therein. Polymers suitable to form a polymeric matrix of a sustained release device of this invention are biocompatible polymers which can be either biodegradable or non-biodegradable polymers, or blends or copolymers thereof.
A polymer, or polymeric matrix, is biocompatible if the polymer, and any degradation products of the polymer, are non-toxic to the recipient and also present no significant deleterious or untoward effects on the recipient~s body, such as an immunological reaction at the injection site.
Biodegradable, as defined herein, means the composition will degrade or erode in vivo to form smaller chemical species. Degradation can result, for example, by enzymatic, chemical and/or physical processes. Suitable CA 02223~83 l997-l2-04 W O 96/40074 PCT~US96/08526 biocompatible, biodegradable polymers include, for example, poly(lactides), poly(glycolides), poly(lactide-co-glycolides), poly(lactic acid)s, poly(glycolic acid)s, poly(lactic acid-co-glycolic acid)s, polycaprolactone, polycarbonates, polyesteramides, polyanhydrides, poly(amino acids), polyorthoesters, polyacetals, polycyanoacrylates, polyetheresters, poly(dioxanone)s, poly(alkylene alkylate)s, copolymers of polyethylene glycol and polyorthoester, biodegradable polyurethanes, blends and copolymers thereof.
Biocompatible, non-biodegradable polymers suitable for a sustained release device include non-biodegradable polymers selected from the group consisting of polyacrylates, polymers of ethylene-vinyl acetates and other acyl substituted cellulose acetates, non-degradable polyurethanes, polystyrenes, polyvinyl chloride, polyvinyl fluoride, poly(vinyl imidazole), chlorosulphonate polyolefins, polyethylene oxide, blends and copolymers thereof.
Further, the terminal functionalities of a polymer can be modified. For example, polyesters can be blocked, unblocked or a blend of blocked and unblocked polyesters.
A blocked polyester is as classically defined in the art, specifically having blocked carboxyl end groups.
Generally, the blocking group is derived from the initiator of the polymerization and is typically an alkyl group. An unblocked polyester is as classically defined in the art, specifically having free carboxyl end groups.
Acceptable molecular weights for polymers used in a sustained release device can be determined by a person of ordinary skill in the art taking into consideration factors such as the desired polymer degradation rate, physical properties such as mechanical strength, and rate of dissolution of polymer in solvent. Typically, an acceptable range of molecular weights is of about 2,000 CA 02223~83 l997-l2-04 Daltons to about 2,000,000 Daltons. In a preferred embodiment, the polymer is a biodegradable polymer or copolymer. In a more preferred embodiment, the polymer is a poly(lactide-co-glycolide) (hereinafter "PLGA") with a lactide:glycolide ratio of about 1:1 and a molecular weight of about 5,000 Daltons to about 70,000 Daltons. In an even more preferred embodiment, the molecular weight of the PLGA
used in the present invention has a molecular weight of about 5,000 Daltons to about 42,000 Daltons.
Typically, a polymeric sustained release microcarrier will contain from about 0.01~ (w/w) to approximately 50~
(w/w) of aggregation-stabilized biologically active agent (dry weight of the composition). The amount of agent used will vary depending upon the desired effect of the agent, the planned release levels, and the time span over which the agent will be released. A preferred range of agent loading is between about 0.1~ (w/w) to about 30~ (w/w) agent. A more preferred range of agent loading is between about 0. 5~ (w/w) to about 20~ (w/w) agent.
In another embodiment, a polymeric sustained release composition also contains a biocompatible metal cation component, which is not contained in the biologically active, aggregation-stabilized particles, but is dispersed within the polymer. The metal cation of this metal cation component acts to modulate the release of the biologically active agent from the polymeric sustained release composition.
This metal cation component typically comprises at least one type of multivalent metal cations. A metal cation component, as defined herein, is a component containing at least one kind of multivalent metal cation (having a valency of +2 or more) in a non-dissociated state, a dissociated state, or a combination of non-dissociated and dissociated states. Suitable metal cation components include, for instance, metal salts, metal CA 02223~83 1997-12-04 WO 96/40074 PCTnUS96/08526 hydroxides, and basic (pH of about 7 or higher) salts of weak acids wherein the salt contains a metal cation. It is preferred that the metal cation be divalent. Examples of metal cation components suitable to modulate release of a biologically active agent, include, or contain, for instance, Mg(OH)2, MgCO3 (such as 4MgCO3 Mg(OH)2 5H2O), ZnCO3 (such as 3Zn(OH)2-2ZnC03), CaCO3, Zn3(C6H5O7)2, Mg(OAc) 2, MgSO4, Zn(OAc) 2~ ZnSO4, ZnCl2, MgCl2 and Mg3(c6H5o7)2. A suitable ratio of metal cation component-to-polymer is between about 1:99 to about 1:2 by weight.
The optimum ratio depends upon the polymer and the metal cation component utilized.
The metal cation component can optionally contain cation species and/or anion species which are contained in an aggregation stabilizer within particles of the agent.
The metal cation component acts to modulate the release of the agent from the polymeric matrix of the sustained release composition and can also enhance the stability of agent in the composition against aggregation. In a modulated release, at least one release characteristic of the agent, such as the initial release level, the subsequent release levels, duration of release and/or the amount of agent released, is different from the release characteristics exhibited by the agent being released from a polymeric matrix, wherein the polymeric matrix does not contain a dispersed metal cation component.
A polymeric matrix containing a dispersed metal cation component to modulate the release of a biologically active agent from the polymeric matrix is further described in co-pending U.S. Patent Application No. 08/237,057, filed May 3, 1994 and co-pending PCT Patent Application PCT/US95/05511, the teachings of which are incorporated herein by reference in their entirety.
In yet another embodiment, at least one pore forming agent, such as a water soluble salt, sugar or amino acid, CA 02223~83 l997-l2-04 is included in a polymeric microparticle to modify the microstructure of the microparticle. The proportion of pore forming agent added to a polymer solution, from which the microparticle is formed, is between about 1~ (w/w) to about 30~ (w/w). It is preferred that at least one pore forming agent be included in a nonbiodegradable polymeric matrix.
The biologically active agent in a sustained release device of the present invention can also contain other excipients, such as stabilizers and bulking agents.
Stabilizers are added to maintain the potency of the biologically active agent over the duration of the agent's release. Suitable stabilizers include, for example, carbohydrates, amino acids, ~atty acids and surfactants and are known to those skilled in the art. For amino acids, fatty acids and carbohydrates, such as sucrose, lactose, mannitol, inulin, maltose, dextran and heparin, the mass ratio of carbohydrate to biologically active agent is typically between about 1:10 and about 20:1. For surfactants, such as polysorbates (e.g., TweenTM) and polyo~Am~rs and polyo~Am;nes (e.g., PluronicTM), the mass ratio of surfactant to agent is typically between about 1:1000 and about 1:20.
Solubility agents can also be added to further modify the solubility of the agent. Suitable solubility agents include complexing agents, such as albumin and protamine, which can be used to slow the release rate of the agent from a polymeric matrix. The weight ratio of solubility agent to biologically active agent is generally between about 1:99 and about 20:1.
Bulking agents typically comprise inert materials.
Suitable bulking agents are known to those skilled in the art.
A polymeric sustained release composition o~ this invention can be ~ormed into many shapes such as a ~ilm, a CA 02223~83 1997-12-04 WO 96/40074 PCTAUS9~'~~Ç~6 pellet, a cylinder, a disc or a microcarrier. A
microcarrier, as defined herein, comprises a polymeric component having a diameter of less than about one millimeter and cont~; n; ng at least one particle of aggregation-stabilized, biologically active agent dispersed therein. A microcarrier can have a spherical, non-spherical or irregular shape. It is preferred that a microcarrier be generally spherical in shape. Typically, the microcarrier will be of a size suitable for injection.
A preferred size range for microcarriers is from about 1 to about 180 microns in diameter, such as for injection through a 23-gauge needle.
In the method for preparing aggregation-stabilized agent, the biologically active agent is mixed with a suitable aggregation-stabilizer. It is understood that either or both the agent and stabilizer can be in solid form, typically particulate, or dissolved in an aqueous solution. It is preferred that the agent and stabilizer be combined in single particles, which are more preferably lyophilized.
In the embodiment wherein a biologically active agent is mixed with a metal cation component to form particles, the agent is mixed in a suitable solvent with at least one suitable metal cation component to form a mixture, wherein each component of the mixture can be in suspension or solution, or a combination thereof. The concentration of agent in solution is typically between about o.l to about 20 mg agent/ml of solvent, and preferentially, between about 1.0 to about 5.0 mg agent/ml of solvent.
In a preferred embodiment, the agent is contacted with at least one suitable aggregation-stabilizing metal cation, such as Ca+2 or Zn+2, and with a suitable solvent, under pH
conditions suitable for forming a complex of the metal cation and the agent. Typically, the complexed agent will be in the form of a cloudy precipitate, which is suspended CA 02223~83 1997-12-04 in the solvent. However, the complexed agent can also be in solution.
~ In an embodiment wherein particles of an agent stabilized with a precipitant are formed, the agent is mixed in a suitable aqueous solvent with at least one suitable precipitant to form a stabilizing mixture, wherein each component of the stabilizing mixture can be in suspension or solution, or a combination thereof.
In forming a stabilizing mixture, the content of precipitant is typically between about 10~ (w/w) and about 80~ (w/w) of the total solids in agent particles and is preferentially more than about 40~ (w/w).
It is understood that the agent can be in a solid or a dissolved state, prior to being contacted with the aggregation stabilizer. It is also understood that the aggregation stabilizer can be in a solid or a dissolved state, prior to being contacted with the agent. In a preferred embodiment, a buffered aqueous solution of an agent is mixed with an aqueous solution of the aggregation stabilizer.
Suitable solvents are those in which the agent and the metal cation component are each at least slightly soluble, such as in an aqueous sodium bicarbonate buffer or in an aqueous phosphate buffer or citrate buffer or combinations thereof. For aqueous solvents, it is preferred that water used be either deionized water or water-for-injection (WFI).
The stabilizing mixture is then dried, such as by lyophilization, to form particulate aggregation-stabilized agent. The stabilizing mixture can be bulk lyophilized or can be divided into smaller volumes which are then lyophilized. In a preferred embodiment, the stabilizing mixture is micronized, such as by use of an ultrasonic ~ nozzle, and then lyophilized to form aggregation-stabilized agent particles. Acceptable means to lyophilize the CA 02223~83 l997-l2-04 WO 96/40074 PCTrUS96J08526 stabilizing mixture include those known in the art. A
solid stabilizing mixture can be pressed into pellets.
A suitable pH range can be achieved by dialysis with a buffer, by using the buffer as a solvent for the agent and/or aggregation stabilizer, and by making one or more bulk additions of buffer to the agent solution before, during, and/or after addition of the aggregation stabilizer.
The stabilizing mixture is usually buffered to a pH
between about 4.0 and about 8.0 to maintain pH in a range which will prevent a significant loss of biological activity resulting from pH changes during particle formation and/or to support formation of complexes. A
preferred pH range is between about 5.0 and about 7.4.
Suitable pH conditions are typically achieved through use of an aqueous buffer, such as sodium bicarbonate, as the solvent for the agent and aggregation stabilizer.
Typically, the content of buffer in a stabilizing mixture is between about 0.1~ (w/w) and about 20~ (w/w) of total solids.
Preferably, particles of aggregation-stabilized agent are between about 1 to about 6 micrometers in diameter.
The agent particles can be fragmented separately, as described in co-pending U.S. Patent Application No. 08/006,682, filed January 21, 1993, which describes a process for producing small particles of biologically active agents, which is incorporated herein in its entirety by reference. Alternately, the agent particles can be fragmented after being added to a polymer solution, such as by means of an ultrasonic probe or ultrasonic nozzle.
The formation of Zn+2-stabilized IFN or hGH particles are further described in Examples 1 and 4.
In one embodiment for forming a sustained release device, a suitable amount of aggregation-stabilized particles of agent is added to a polymer solution. The CA 02223~83 1997-12-04 W O 96/40074 PCTrUS96/08526 agent particles can be dispersed with the polymer solution by stirring, agitation, sonication or by other known mixing means. The polymer solution, having a dispersion of biologically active, aggregation-stabilized agent is then solidified, by appropriate means, to form a sustained release composition of this invention.
Alternately, biologically active, aggregation-stabilized particles of agent and a polymer can be mixed into a polymer solvent sequentially, in reverse order, intermittently, separately or through concurrent additions, to form a dispersion of the agent particles in a polymer solution.
A suitable polymer solution contains between about 1~ (w/w) and about 30~ (w/w) of a suitable biocompatible polymer, wherein the biocompatible polymer is typically dissolved in a suitable polymer solvent. Pre~erably, a polymer solution contains about 2~ (w/w) to about 20~ (w/w) polymer. A polymer solution containing about 5~ to about 15~ (w/w) polymer is most preferred.
A suitable polymer solvent, as defined herein, is solvent in which the polymer is soluble aggregation-stabilized particles of agent are substantially insoluble and non-reactive. Examples of suitable polymer solvents include polar organic liquids, such as methylene chloride, chloroform, ethyl acetate, acetone methylisobutylketone, n-butylacetate, isobutyl acetate, tetrahydrofuran, methyl acetate and ethyl citrate.
In yet another embodiment of the method of this invention, a metal cation component, not contained in the aggregation-stabilized particles of biologically active agent, is also dispersed within the polymer solution to modulate the release of the biologically active agent.
It is understood that a metal cation component and the aggregation-stabilized particles can be dispersed into a polymer solution sequentially, in reverse order, CA 02223~83 1997-12-04 WO 96/40074 PCT~US96/08526 intermittently, separately or through concurrent additions.
Alternately, a polymer, a metal cation component and the aggregation-stabilized particles can be mixed into a polymer solvent sequentially, in reverse order, intermittently, separately or through concurrent additions.
The method for forming a composition for modulating the release of a biologically active agent from a biodegradable polymer is further described in co-pending U.S. Patent Application No. 08/237,057 and co-pending PCT
lG Patent Application PCT/US95/05511.
One suitable method for forming a sustained release composition from a polymer solution is the solvent evaporation method described in U.S. Patent No. 3,737,337, issued to Schnoring et al., U.S. Patent No. 3,523,906, issued to Vr~nchen et al., U.S. Patent No. 3,691,090, issued to Kitajima et al., or U.S. Patent No. 4,389,330, issued to Tice et al. Solvent evaporation can be used as a method to form microcarriers and other shaped sustained release devices.
In the solvent evaporation method, a polymer solution containing a dispersion of particles of an aggregation-stabilized biologically active agent, is mixed in or agitated with a continuous phase, in which the polymer solvent is partially miscible, to form an emulsion. The continuous phase is usually an aqueous solvent.
Emulsifiers are often included in the continuous phase to stabilize the emulsion. The polymer solvent is then evaporated over a period of several hours or more, thereby solidifying the polymer to form a polymeric matrix having a dispersion of particles of aggregation-stabilized biologically active agent contained therein.
In this method, care must be taken not to heat the polymer solution to a temperature at which denaturing of the biologically active agent in the aggregation-stabilized particles could occur.

CA 02223~83 1997-12-04 WO 96/40074 PCT/U~,G,~ 6 Another suitable method for solidifying a polymer solution to form a polymeric matrix, containing particles of aggregation-stabilized biologically active agent, is the phase separation method described in U.S. Patent No.
4,675,800, which is incorporated herein in its entirety by reference. In this method, polymer within a polymer solution cont~;n;ng aggregation-stabilized particles is precipitated around the particles by the addition of a polymer non-solvent to the polymer solution to form an emulsion, wherein the polymer non-solvent is immiscible with the polymer solvent.
A preferred method for forming aggregation-stabilized microcarriers from a polymer solution uses rapid freezing and solvent extraction as described in U.S. Patent No.
5,019,400, issued to Gombotz et al. and co-pending U.S.
Patent Application No. 08/433,726, filed May 18, 1995, the teachings of which are incorporated herein in their entirety by reference. This method of microcarrier formation, as compared to other methods, such as phase separation, additionally reduces the amount o~ biologically active agent required to produce a sustained release composition with a specific content and also minimizes the loss of biological activity during microparticle formation.
Further discussion of the high level of biological activity, typically ~ 98~, maintained in the microparticles of the present invention, formed using this preferred method, is provided in Example 2. Also see Examples 2, 5 and 7 for additional descriptions of microparticle formulations by this method.
In this method, the polymer solution, containing the dispersion of aggregation-stabilized particles, is processed to create droplets, wherein at least a significant portion of the droplets contain polymer solution and aggregation-stabilized particles. These droplets are then frozen by means suitable to form CA 02223~83 1997-12-04 W O 96140074 PCT~US96/08526 microparticles. Examples of means for processing the polymer solution dispersion to form droplets include directing the dispersion through an ultrasonic nozzle, pressure nozzle, Rayleigh jet, or by other known means for creating droplets ~rom a solution.
Means suitable ~or ~reezing droplets to form microparticles include directing the droplets into or near a liquified gas, such as liquid argon and liquid nitrogen to form frozen microdroplets which are then separated from the liquid gas. The frozen microdroplets are then exposed to a liquid non-solvent, such as ethanol, or ethanol mixed with hexane or pentane. The solvent in the ~rozen microdroplets is extracted as a solid and/or liquid into the non-solvent to form microcarriers containing aggregation-stabilized biologically active agent. Mixing ethanol with other non-solvents, such as hexane or pentane, can increase the rate o~ solvent extraction, above that achieved by ethanol alone, from certain polymers, such as poly(lactide-co-glycolide) polymers.
A wide range of sizes of sustained release microcarriers can be made by varying the droplet size, for example, by changing the ultrasonic nozzle diameter. I~
very large microcarriers are desired, the microcarriers can be extruded through a syringe directly into the cold liquid. Increasing the viscosity of the polymer solution can also increase microparticle size. For example, the size of the microcarriers produced by this process can vary over a wide range, such as from greater than about 1000 to about 1 micrometers, or less, in diameter.
Yet another method of forming a sustained release composition, from a polymer solution, includes film casting, such as in a mold, to form a film or a shape. For instance, a~ter putting the polymer solution containing a dispersion of aggregation-stabilized particles into a mold, the polymer solvent is then removed by means known in the CA 02223~83 1997-12-04 W O 96/40074 PCT~US96/08526 art, or the temperature of the polymer solution is reduced, until a film or shape, with a consistent dry weight, is obtained. Film casting of a polymer solution, containing a biologically active agent, is further described in co-pending U.S. Patent Application No. 08/237,057.
It is believed that the release of the biologically active agent can occur by two different mechanisms. The agent can be released by diffusion through aqueous filled ~h~nnels generated in the polymeric matrix, such as by the dissolution of the agent or by voids created by the removal of the polymer's solvent during the synthesis of the sustained release composition. A second mechanism is the release of the agent due to degradation of the polymer.
The rate of polymer degradation can be controlled by changing polymer properties that influence the rate of hydration of the polymer. These properties include, for instance, the ratio of different monomers, such as lactide and glycolide, comprising a polymer; the use of ~he L-isomer of a monomer instead of a racemic mixture; the polymer end group; and the molecular weight of the polymer.
These properties can affect hydrophilicity and crystallinity, which control the rate of hydration of the polymer. Hydrophilic excipients such as salts, carbohydrates and surfactants can also be incorporated to increase hydration and which can alter the rate of erosion of the polymer.
By altering the properties of the polymer, the contributions of diffusion and/or polymer degradation to the release of biologically active agent can be controlled.
For example, increasing the glycolide content of a - poly(lactide-co-glycolide) polymer and decreasing the molecular weight of the polymer can enhance the hydrolysis - of the polymer and thus, provides an increased agent release from polymer erosion.

CA 02223~83 1997-12-04 In addition, the rate of polymer hydrolysis may be increased in non-neutral pH's. Therefore, an acidic or a ba~ic excipient can be added to the polymer solution, used to form the microcarriers, to alter the polymer erosion rate.
The sustained release device of this invention can be administered to a human, or other ~n;m~l, by injection, implantation (e.g, subcutaneously, intramuscularly, intraperitoneally, intracranially, intravaginally and intradermally), ~m; n; stration to mucosal membranes (e.g., intranasally or by means of a suppository), or in si tu delivery (e.g. by enema or aerosol spray) to provide the desired dosage of an agent based on the known parameters for treatment with that agent of the various medical conditions.
The invention will now be further and specifically described by the following examples.

Example 1 Formation of Aqqreaation-Stabilized Interferon IFN-~,2b, which was used in the present Examples, is identical to IFN-~,2 as described in Rubenstein et al ., Biochem. Biophys. Acta, 695: 705-716 (1982), with the exception that the lysine at position 23 of IFN-~,2 is an arginine in IFN-~,2b. The IFN was-stabilized by forming a complex with zn+2 ions, wherein the complex has a lower solubility in aqueous solutions than does non-complexed IFN.
The IFN was complexed as follows. The IFN-~,2b was dissolved in different volumes of 10 mM sodium bicarbonate buffer (pH 7.2) to form IFN solutions with concentrations between 0.1 and 0.5 mM IFN. A 10 mM zn+2 solution was prepared from deionized water and zinc acetate dihydrate and then was added to the IFN solutions to form Zn+2-IFN
solutions with a final IFN concentration of about 1.3 mg/ml CA 02223~83 l997-l2-04 and a Zn+2: IFN molar ratio of 2:1, 4:1 or 10:1, respectively. The pH of the Zn+2- IFN solution was then adjusted to 7.1 by adding 1~ acetic acid. A cloudy suspended precipitate, comprising aggregation-stabilized IFN wherein the IFN iS stabilized as a complex with Zn+2, formed in each solution.
The suspension of aggregation-stabilized IFN was then micronized using an ultrasonic nozzle (Type VlA; Sonics and Materials, Danbury, CT) and sprayed into a polypropylene tub (17 cm diameter and 8 cm deep) containing liquid nitrogen to form frozen particles. The polypropylene tub was then placed into a -80 ~C freezer until the liquid nitrogen evaporated. The frozen particles, which contained Zn+2-stabilized IFN, were then lyophilized to form aggregation-stabilized IFN particles.

Exam~le 2 Pre~aration of PLGA Microcarriers Containinq Aqqreqation-Stabilized IFN
Samples of blocked PLGA (intrinsic viscosity of 2C 0.15 dl/g) obtained from Birmingham Polymers (Birmingham, AL) or a hydrophilic unblocked PLGA ( intrinsic viscosity of 0.17 dl/g) obtained ~rom Boehringer Ingelheim Chemicals, Inc. (Montvale, NJ), were dissolved in 10 ml of methylene chloride per gram of PLGA to form polymer solutions. To these polymer solutions were added about 0.033, 0.1 or 0. 2 grams of aggregation-stabilized IFN particles per gram of PLGA, formed as described in Example 1 to form polymer solutions with the following formulations:

CA 02223~83 l997-l2-04 WO 96/40074 PCT~US96/08526 Zn:IFN IFN:PLGA IFN:MgCO3 IFN:ZnCO3 Formula PLGA Molar Ratio Mass Ratio Mass Ratio Mass Ratio 1 Blocked 2:1 0.2:1 N/A N/A
2 Blocked 4:1 0.2:1 N/A N/A

3 Blocked 10:1 0.2:1 N/A N/A

4 Blocked 2:1 0.1:1 1:1 N/A

5 Unblocked 2:1 0.033:1 1:1 N/A

6 Blocked 2:1 0.033:1 N/A 3:1 7 Blocked 2:1 0.1:1 N/A 1:1 8 Blocked 2:1 0.1:1 N/A 8:1 When added to the polymer solutionL MgCO3 and ZnCO3 were sieved through a 38 micrometer (#400) sieve. Each formulation was then sonicated using an ultrasonic probe (Virtis, Co., Gardiner, NY) to ~ragment and suspend aggregation-stabilized IFN particles in the polymer solutions. The size of the sonicated, aggregation-stabilized IFN particles was between about 2-15 microns.
The suspension was then placed in a 10 ml gas-tight syringe.
About 400 ml o~ 100~ ethanol per gram PLGA was added to a round polypropylene tub. This solution was frozen by surrounding the tub with liquid nitrogen. The ~rozen ethanol was then covered with 500 ml o~ liquid nitrogen per gram o~ PLGA. The IFN suspension was then pumped ~rom the syringe by a syringe pump (Orion Sage Pump Model 355, Orion Research Inc., Boston, MA), at a rate o~ 1.7 ml/min, into an ultrasonic nozzle (Type VlA, Sonics and Materials, Danbury, CT) that was placed above the container containing the ~rozen ethanol covered with liquid nitrogen. The nozzle atomized the IFN suspension into droplets which ~roze upon contact with the liquid nitrogen and ~ormed microcarriers which sank to the sur~ace o~ the ~rozen ethanol.

CA 02223~83 1997-12-04 The container was placed into a -80 ~C freezer, thereby evaporating the liquid nitrogen and allowing the ethanol to melt. As the ethanol thawed, the microcarriers sank into it. The temperature was lowered to -95.1 ~C and the methylene chloride was extracted from the microcarriers. After 24 hours, an additional 400 ml of 100~ ethanol per gram of PLGA, which was prechilled to -80 ~C, was added to the container. Three days after the microcarriers were prepared, the ethanol/microcarrier slurry was filtered using a 0.65 micron DuraporeTM membrane (Millipore, Bedford, MA). The filtered microcarriers were then vacuum dried in a lyophilizer.

Exam~le 3 In vitro Release of IFN Enca~sulated with Non-Metal Cation Stabilizer ComPared to IFN Stabilized with zn+2 Dextran 70 (Spectrum Chemical Manu~acturing Co., Gardena, CA) was added to a solution of IFN-~,2b in 10 mM
sodium phosphate buffer at a weight ratio of 1:1 (Dextran:IFN). The solution was micronized through an ultrasonic nozzle as described in Example 1 and the frozen particles were then lyophilized. The IFN-Dextran particles were subsequently microencapsulated in blocked PLGA as described in Example 2 to form IFN-Dextran microcarriers.
Aggregation-stabilized IFN particles (2:1 Zn+2:IFN ratio), as described in Example 1, were also microencapsulated as described in Example 2 to form aggregation-stabilized IFN
microcarriers.
In vitro dissolution was conducted on the two microcarrier formulations by incubating 20 mg of each type of microcarrier in buffer at 37 ~C. IFN release from the microcarriers was monitored by BioRad protein assay (BioRad Inc. Richmond, CA).

CA 02223~83 1997-12-04 IFN release from the IFN-Dextran microcarriers was linear for the first 10 days with an average release rate of 6.4~/day. The release continued at a rate of 0.4~/day from day 10 to day 14 with a total cumulative release of 66~ by day 14. No further release of protein from the microcarriers was detected. The microcarriers were dried down at day 28. The IFN-Dextran r~m~;n;ng was extracted from the microcarriers and the protein was characterized by testing its solubility in water and monomer content by sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE). Only 18~ of the protein remaining inside the microcarriers was water soluble. The insoluble protein was solubilized using SDS and run on a gel. The insoluble material contained 19~ covalent aggregates and 81~ non-covalent aggregates.
In contrast the microcarriers with the IFN
aggregation-stabilized with zn+2 showed linear release for at least 28 days at a rate of 2.7~/day. The analyses indicate the formulation of IFN with zinc is more stable resulting in a longer period o~ continuous release o~
protein from the microcarriers.

Exam~le 4 Formation of Aqqreqation-Stabilized hGH
Purified recombinant human growth hormone (hGH), whose DNA sequence is as described in U.S. Patent 4,898,830, issued to Goeddel et al., was used in this Example. The human growth hormone was stabilized by forming a complex with zn+2 ions, wherein the complex has a lower solubility in aqueous solutions than does non-complexed hGH.
The hGH was dissolved in samples of a 4 mM sodium bicarbonate buffer (pH 7.2) to form hGH solutions with concentrations between 0.1 and 0.5 mM hGH. A 0.9 mM zn+2 solution was prepared from deionized water and zinc acetate dihydrate and then was added to the hGH solutions to ~orm CA 02223~83 l997-l2-04 W O 96/40074 PCT~US96/08526 Zn+2-hGH solution. The pH of the Zn+2-hGH solution was then adjusted to between 7.0 and 7.4 by adding 1~ acetic acid. A cloudy suspended precipitate, comprising Zn+2-stabilized hGH formed. Lyophilized, aggregation-stabilized hGH particles were then formed as described in Example 1.

Example 5 Pre~aration and Analysis of PLGA Microcarriers Containing BioloqicallY Active, Aaareaation-Stabilized hGH
Microcarriers containing aggregation-stabilized hGH, ~ormed as described in Example 4, were prepared using the method of Example 2 from hydrophilic unblocked PLGA (50:50 PLGA, 9,300 Daltons; RG502H polymer; Boehringer Ingelheim Chemicals, Inc.), blocked PLGA t50:50 PhGA, 10,000 Daltons;
Lot #115-56-1, Birmingham Polymers, Inc., Birmingham, AL) and unblocked PLGA (50:50 PLGA, 31,000 Daltons; RG503H, Boehringer Ingelheim Chemicals, Inc.) and varying amounts oi~ ZnC03.
The integrity o~ the hGH encapsulated in microcarriers was determined by extracting the hGH ~rom the microcarriers. The microcarriers were placed in a tube containing methylene chloride and vortexed at room temperature to dissolve the polymer. Acetone was then added to the tube, which was subsequently vortexed, to extract and collect the hGH. The collected hGH was then ~reeze-dried and re-constituted in HEPES bu~er containing 10 mM EDTA. Appropriate controls were run to ensure that the extraction process did not a~fect the integrity o~ the protein.
The integrity o~ the encapsulated hGH was analyzed by measuring the percent o~ hGH monomer contained in the hGH
sample after encapsulation by size exclusion chromatography (SEC).

CA 02223~83 1997-12-04 WO 96/40074 PCT~U~9G/ RC~6 The results of SEC analyses of the hGH integrity of hGH sustained release microcarriers were provided below.

% Zinc Carbonate)~ Monomer (SEC) 31K unblocked; 6~ ZnC0398.6 31K unblocked; 6~ ZnC0399.2 31K unblocked; 3~ ZnC0397.7 31K unblocked; 3~ ZnC0397.8 31K unblocked; 1~ ZnC0397.6 31K unblocked; 0~ ZnC0397.8 31K unblocked; 0~ ZnC0397.1 lOK blocked; 1~ ZnC03 98.2 lOK blocked; 1~ ZnC03 98.4 8K unblocked; 0~ ZnC03 98.5 lOK blocked; 1~ ZnC03 98.4 The results showed that the encapsulation process did not cause aggregation of the protein.

Exam~le 6 Formation of Aaqreqation-Stabilized EP0 Erythropoietin was derived as described in U.S. Patent No. 4,703,008. The EP0 was dissolved in deionized water to form an aqueous solution having a concentration of approximately 1 mg/ml. Different samples of the EP0 solution were then dialyzed against three changes of the appropriate formulation buffer (i.e., 5mM phosphate buffer (pH 7), 5 mM citrate buffer (pH 7), 5 mM citrate/5mM
phosphate buffer (pH 7) or 10 mM bicarbonate buffer (pH 7)).
Following dialysis, the concentration of EP0 in the~0 dialyzed solutions was verified to be approximately 1 mg/ml CA 02223~83 1997-12-04 as determined by measuring absorbance at 280 nm (~ =
1.345 L gm~1 cm~1).
Portions of the dialyzed EPO solutions were then mixed with concentrated solutions of candidate anti-aggregation agents (i.e., ~mm~;um sulfate, zinc acetate, mannitol/sucrose or mannitol/maltose) to form the EPO
formulations provided in Table I below. The candidate anti-aggregation agent solutions also possibly contained additional excipients (i.e, inulin, glycine and TWEEN 20 surfactant).
The anti-aggregation agent solutions were separately prepared in the same buf~ers used to dialyze the EPO
solutions to which they were subsequently added.
Approximate volumes of each anti-aggregation agent solution and of additional buffer were added to a 50 ml polypropylene tube to achieve the desired concentrations for the formulations (described in Table I). Each dialyzed EPO solution was then added to the appropriate anti-aggregation agent solution and then the solutions were mixed by gentle inversion.

CA 02223~83 1997-12-04 W O 96/40074 PCT/U'~-'0~'~6 Table I

Formulations Aml Am4 Am7 Mal Ma3 Ma4 Znl Zn6 (wt ~) EPO 10.0 10.1 9.9 10.0 10.0 10.0 10.0 10.0 5 Ammonium Sul~ate 66.8 64.7 79.1 0.0 0.0 0.0 0.0 0.0 Zinc Acetate 0.0 0.0 0.0 0.0 0.0 0.0 76.9 76.9 Mannitol 0.0 0.0 0.0 62.5 62.5 72.5 0.0 0.0 Sucrose 0.0 0.0 0.0 10.0 0.0 10.0 0.0 0.0 Maltose 0.0 0.0 0.0 0.0 10.0 0.0 0.0 0.0 5mM Citrate Buffer (pH 7) 0.0 15.1 0.0 0.0 0.0 0.0 0.0 0.0 5mM Phosphate Buffer (pH 7) 0.0 0.0 10.0 7.5 7.5 7.5 0.0 0.0 15 5mM Citrate/
5mM Phosphate Buffer (pH 7) 22.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 lOmM
Bicarbonate Buffer (pH 7) 0 0 0.0 0.0 0.0 0.0 0.0 13.1 12.1 Inulin 1.1 10.1 1.0 0.0 0.0 0.0 0.0 0.0 Glycine 0.0 0.0 0.0 10.0 10.0 0.0 0.0 0.0 Sur~actant 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 Lyophilized, aggregation-stabilized EPO particles were then ~ormed from the EPO solutions as described in Example 1. The EPO particles were removed from the lyophilizer under an atmosphere of dry nitrogen, handled in a low humidity environment, and stored desiccated at -80~C.

CA 02223~83 1997-12-04 Exam~le 7 PreParation and Analysis of PLGA Microcarriers Containinq Agqreqation-Stabilized Erythro~oietin Microcarriers containing the aggregation-stabilized EPO formulations of Example 6 were prepared from unblocked (50:50; MW 10,000 Daltons) PLGA, obtained from Boehringer Ingelheim Chemicals, Inc., Montvale, NJ, or blocked (50:50;
MW 10,000 Daltons) PLGA obtained from Birmingham Polymers, Inc., Birmingham, AL.
In addition, microcarriers, containing the Am7 formulation of aggregation-stabilized EPO particles, were prepared ~rom unblocked (50:50) PLGA with a molecular weight of approximately 31,000 Daltons or 45,000 Daltons, (Boehringer Ingelheim Chemicals, Inc., Montvale, NJ).
The method described in Gombotz et al. (U.S. Patent No. 5,019,400), and in Example 2, was used to encapsulate the aggregation-stabilized EPO particles of Example 6 in PLGA. In each case, polymer was dissolved in 5.1 ml of methylene chloride to form a polymer solution. Magnesium carbonate, or zinc carbonate, was sieved through a 38 micrometer sieve and was then added to the polymer solution to a final concentration of 10~ w/vol. The polymer/salt suspension was subsequently combined with 30 mg of aggregation-stabilized EPO particles.
The polymer solution, containing suspended salt and EPO particles, was placed in an ice-water bath and sonicated using an ultrasonic probe (Virtis Co., Gardiner, NY) to reduce the protein particle size to approximately 2 to 3 micrometers in diameter and to ~orm a dispersion of EPO particles within the polymer solution.
Microcarriers containing aggregation-stabilized EPO
were prepared using the method described in Example 2.
The ;mmllnoreactivity of the EPO in these sustained release microcarriers was subsequently determined by extracting protein and analyzing by radio;mmllnoassay (RIA) CA 02223~83 1997-12-04 W O 96/40074 PCT~US9~/0~3~6 (Incstar: Stillwater, MN). To extract the EPO from the microcarriers, approximately 10 mg of microcarriers were placed in a tube with 250 ~l of methylene chloride. The samples were vortexed for 10 to 20 seconds and left at room temperature for 5 minutes to dissolve the polymer. A
sample of acetone (750 ~l) was added, vortexed for an additional 10 seconds, and centrifuged at 14,000 rpm for 30 seconds at 4 ~C to pellet the EPO. The supernatant was removed and the methylene chloride and acetone steps were repeated twice more. Samples were dried in a lyophilizer or vacuum oven for 14-18 hours at room temperature. The EPO pellet was reconstituted in 1 ml HEPES buffer by vortexing for about 10 seconds, then standing at room temperature for about 1 hour until completely dissolved.
The extracted EPO was diluted in buffer (8.1 mM Na2HPO4, 1.5 mM KH2PO4, 400 mM NaCl, pH 7.5) to a concentration of approximately 25 ~g/ml for analysis.
The ;mml~noreactivity of the EPO was found to be 121,000 + 5000 units per mg of EPO. This specific activity is comparable to the range obtained for bulk EPO (130,000-140,000 units per mg of EPO) thus showing an insignificant reduction of EP0 activity due to the method of forming the sustained release compositions of the present invention.
Monomer content was found to be greater than 98~ for all microcarriers.
The microcarriers containing Aml and Am7 EPO particles were also assayed for EPO dimer, by size exclusion chromatography (SEC), and for high molecular weight EPO
aggregates by SDS-PAGE/ Western blot analysis. No EPO
dimer or high molecular weight aggregates were detected.

CA 02223~83 1997-12-04 W O 96/40074 PCT~US96/08526 Exam~le 8 In Vitro Release of EPO From Aqqreaation-Stabilized BPO Microcarriers The in vitro release kinetics of EPO from aggregation-stabilized particles within PLGA microcarriers wereassessed in HEPES buffer (75 mM HEPES, 115 mM NaCl, 0.1 (by volume) TWEEN 20~, 0.1~ (by weight) sodium azide titrated to pH 7.4 with NaOH) or in HEPES buffer containing 2~ or 20~ sheep serum. The studies were conducted by suspending 8-10 mg of microcarriers in 1-5 ml of buffer at 37~C. At specified time points, the buffer was removed in full and replaced with fresh buffer.
In samples incubated in HEPES buffer, the releases over time of EPO monomer (biologically active EPO) and of EPO aggregates (biologically inactive EPO) were determined by size exclusion chromatography (SEC). The results of the SEC analyses upon in vitro release kinetics in HEPES buffer of various microcarriers, wherein the microcarriers were a) unblocked PLGA (MW 10,000 Daltons) microcarriers containing formulations Aml or Am7, and b) blocked PLGA (MW 10,000 Daltons) microcarriers containing Znl, are provided in Figures 1, 2 and 3, respectively. Figures 1 and 2 show the EPO released from formulations containing ~mmon; um sulfate as an anti-aggregation agent was almost all monomeric EPO
over the length of the release period.
Figure 3 shows the EPO released from a formulation containing zinc acetate, as an anti-aggregation agent, contained significant levels of aggregate which increased substantially over the length of the release period.
The results of the SEC and RIA analyses upon in vi tro release kinetics in HEPES buffer, and in HEPES/serum, of various microcarriers (all in 10,000 Dalton PLGA) which contained different EPO formulations of Example 6 are provided in Table II. The initial burst and release rate were determined in the HEPES/serum test by RIA. The CA 02223~83 1997-12-04 integrity of the released EPO was assessed in HEPES buffer by SEC.

Table II
Formula EPO PolYmer/ Aqareqate Initial Averaqe Release Load Salt Released BurstRelease Duration ~ init.load) (~/day) (days) Znl 10 Blocked/12 66 1.2 14 10~ MgCO3 Znl 10 Blocked/22 46 1.7 28 10~ ZnC03 Zn6 10 Blocked/37 32 1.6 28 10~ ZnC03 Aml 5 Unblocked/ 1 39 1.4 21 10~ MgCO3 Aml 10 Blocked/ 2 71 0.3 3 10~ MgCO3 Am4 5 Unblocked/ 1 29 1.1 21 10~ MgCO3 Am4 5 Unblocked/ 1 35 0.9 28 none Mal 5 Unblocked/ 1 44 1.8 24 10~ MgCO3 Ma3 10 Unblocked/ 1 71 1.3 21 10~ MgCO3 Ma4 10 Blocked/ 1 77 0.6 3 10~ ZnC03 These analyses show that the addition of suitable anti-aggregation agents significantly reduced the aggregation of EPO over the release periods. These analyses also demonstrated that the addition of a metal cation component (e.g., salt) to the polymer, as well as the selection of the type of polymer (i.e., blocked or unblocked) significantly affected the initial burst level and the duration of release.

CA 02223~83 l997-l2-04 Exam~le 9 Inteqrity of EPO Released In Vitro from Aqqreqation-Stabilized EPO Microcarriers The purpose of the experiment was to determine the integrity of EPO released from PLGA microcarriers having varying concentrations of ammonium sulfate.
Aggregation-stabilized EPO formulations comparable to Am7, except having 10~, 20~, or 40~ ammonium sulfate, were prepared as described in Example 6. The eliminated ammonium sulfate was replaced with sodium chloride or sucrose such that the total weight of ammonium sulfate and sodium chloride or sucrose was 79~.
The percent monomeric and aggregate EPO were determined after 35 days and 42 days release in vitro. The Am7 formulation, as well as the 40~ ammonium sulfate/NaCl formulation produced 3-4~ aggregates at both time points, whereas the 10~ and 20~ ~mm~n; um sulfate/NaCl formulations produced 5-6~ aggregates. Mannitol formulations produced results similar to the 10% and 20~ ammonium sulfate formulations.
In the case where ammonium sulfate was replaced with sucrose, there was not sufficient drug released from the 40~ ammonium sulfate formulation to auantitate. The 10%
and 20~ ammonium sulfate formulations with sucrose, like their sodium chloride counterparts, showed more aggregates (6-9~) than were observed with the Am7 formulation.

Exam~le 10 In Vivo Release of Aaqreqation-Stabilized IFN-~,2b From Polymeric Microcarriers in Rats Microcarriers, containing aggregation-stabilized IFN, which were prepared as described in Example 2, were tested in rats for the in vivo release of IFN-~,2b. Normal rats were obtained from Taconics, Inc. (Germantown, New York).
The ~n;m~l s were fed with a standard diet and allowed ~ree CA 02223~83 1997-12-04 WO 96/40074 PCT/U'~ 5 access to water. Three to four rats were injected subcutaneously in the interscapular region with a dose of 0.6 - 2.0 mg of IFN/kg, in a 0.5~ gelatin, 1~ glycerol and 0.9~ w/w NaCl vehicle, on day 0 for each of the IFN
microcarriers of Example 2. Blood samples were taken from the tail vein o~ each rat at 1, 2, 4, 8, 10 (optionally), 24, 36 and 48 hours after injection. Additional blood samples were then taken approximately once a day for the following 4-5 days. The IFN concentration in the rat serum samples was determined using an IFN-~ immunoradiometric assay, (Celltech, Slough, U.K), hereinafter "IRMA". The IRMA assay has a m;n;mllm limit of detecting of 6 IU/ml.
The IFN serum levels for control rats, which did not receive the microcarriers containing Zn+2-stabilized IFN
were found to be less than 6 IU/ml.
The results of the IRMA assays conducted on the rats receiving the microcarriers of Example 2 are shown in Figures 4-10. Figures 4-10 show that these injectable microcarrier formulations provided a sustained release of immunologically active IFN-~.

Example 11 In Vivo Release of Aqqreqation-Stabilized IFN
From Polvmeric Microcarriers in ImmunosuPpressed Rats One group of male Sprague-Dawley rats (N=2) (control group), weighing 400 + 50 g (S.D.) was injected as described in Example 10 with the microcarriers of Formula 8 of Example 2. An additional group (N=2) of rats (test group) was also given daily intraperitoneal injections of 10 mg cyclosporin A (Sandimmune~ Injection, Sandoz, East Hanover, NJ) and 5 mg hydrocortisone (Spectrum Co., Gardena, CA) in 0.5 ml sterilized saline for injection (USP) per Kg of body weight for days 0 to 14 and then injections twice a week for days 15 to 28. These injections were to suppress the response of the rats' CA 02223~83 1997-12-04 W O 96/40074 PCT/U~ C~6 immune systems to the release of IFN-~,2b in vivo. No antibody titers were detected in these rats for the duration of treatment.
This method of ;mml~nosuppression is ~urther described in co-pending U.S. Patent Application No. 08/480,813, ~iled June 7, 1995.
The control group did not receive injections to suppress their immune response to IFN-~,2b. Antibodies were detected after day 7 in these rats.
The serum levels of IFN-~,2b in the rats of the experimental group and the control group were determined by IRMA through day 29 (696 hours and 480 hours, respectively). These results are provided in Figure 11.
The results for both groups are the same through day 7 suggesting that the cyclosporin A/hydrocortisone treatment does not affect the measured serum concentrationc of IFN.
The results show that the control group serum levels measured for IFN were artificially high due to their production of antibodies to the IFN-~,2b. The results for the experimental group, in which antibody ~ormation was suppressed, showed sustained release of IFN-~,2b for up to at least 29 days for the pre~erred microcarriers (Formula 8) of Example 2.

Example 12 In Vivo Release of IFN-~,2b ~rom Aaqreqation-Stabilized IFN Microcarrier in Monkeys Microcarriers (Formula 8), as prepared as in Example 2, were tested in a test group of four male cynomolgous monkeys (Charles River Primates) for release o~~ 30 IFN-~,2b. The ~n;m~ S were fed with a standard diet and allowed free access to water. Each monkey was injected subcutaneously with a dose of about 0.12 mg IFN/kg monkey on day zero.

CA 02223~83 1997-12-04 WO 96140074 PCT~US9~ 5 Concurrently, each monkey in a control group of four monkeys, with the same diet and water access as the test group, were subcutaneously injected with an aqueous saline solution containing about 0.12 mg IFN/kg monkey.
Blood samples were taken from the femoral vein at 0, 1, 3, 6, 12, 24, 48, 96, 120, 144, 168, 240, and 336 hours after injection. The IFN-~,2b concentration in the monkey serum samples was determined using both a cytopathic effect assay (CPE; Pharmacopeial Previews, United States Convention, Inc., Nov-Dec 1990, page 1241) and IRMA. The CPE results for both groups are provided in Figure 12.
For the test group, the IRMA and CPE results were similar and showed sustained release of IFN-~,2b from the microcarriers.
The CPE and IRMA results for the control group, which received the aqueous IFN-~,2b injection, showed that the IFN-~,2b concentration fell below detectable limits before the second day of testing.
Figure 12 shows that the microcarrier formulation injected provided sustained release of biologically active IFN-~.

Exam~le 13 Assay for hGH After in Vivo Deqradation of Aqqreqation-Stabilized hGH Microcarriers Microcarriers of blocked-PLGA, containing 15~ w/w Zn+2-stabilized hGH and 0~, 6~, 10~ or 20~ ZnCO3 were formed by the method of Example 5. Groups of test rats were injected subcutaneously with 50 mg samples of the different hGH microcarriers. The rats were sacrificed after 60 days and the skin samples were excised from the injection sites. The excised skin samples were placed in 10~ Neutral Buffered Formalin for at least 24 hours. They were then trimmed with a razor blade to remove excess skin and placed in PBS.

CA 02223~83 1997-12-04 WO 96/40074 PCT/U~,6/~~C~6 Tissue samples were processed by Pathology Associates, Inc. (Frederick, MD). The skin samples were embedded in glycomethacrylate, sectioned and assayed for the presence of hGH using a HistoScan/LymphoScan Staining Kit (Product #24-408M; Accurate Chemical ~ Scientific Corp., Westbury, NY) according to the manufacturer's instructions. Tissue samples were scored ~or the presence or absence of staining which was indicative of the presence or absence of hGH in the sample.
All skin samples, associated with hGH microcarrier injections, tested positive for the presence of hGH thus indicating that the blocked-PLGA microcarriers still contained hGH after 60 days in vivo.
The method described in Example 5 was used to form microcarriers by encapsulating 0% or 15~ w/w hGH, in the form of Zn:hGH complex, and also 0~ or 6~ w/w ZnCO3 salt, within blocked-PLGA and within unblocked-PLGA.
In vivo degradation of unblocked-PLGA microcarriers versus blocked-PLGA microcarriers were compared by injecting samples of microcarriers into rats and then analyzing the microcarriers r~;n;ng at the injection site at various times post-injection. Three rats were assayed at each time point for each microcarrier sample. On the day o~ administration of the microcarriers, 750 ~l o~
vehicle (3~ carboxymethyl cellulose (low viscosity) and 1~
Tween-20 in saline) was added to vials containing 50 + 1 mg of microcarriers. Immediately, the vials were shaken vigorously to form a suspension which was then aspirated into a 1.0 cc syringe without a needle.
Rats (Sprague-Dawley males) were anesthetized with a halothane and oxygen mixture. The injection sites (intrascapular region) were shaven and marked with a permanent tatoo to provide for the precise excision of skin at the sampling time points. Each rat was injected with an entire vial of microcarriers using 18 to 21 gauge needles.

CA 02223~83 1997-12-04 W 096/40074 PCT~US96/08526 On designated days (days 15, 30, 59 and 90 post-injection for ~n;m~l S receiving blocked-PLGA microcarriers, or days 7, 14, 21, 28 and 45 post-injection for animals receiving unblocked-PLGA microcarriers) the rats were sacrificed by asphyxiation with CO2 gas and the skin at the injection sites (including microcarriers) was excised.
Since the microcarriers tended to clump at the injection sites, the presence or absence of microcarriers was determined visually.
The visual inspections found that the unblocked-PLGA
microcarriers degraded substantially faster than the blocked-PLGA microcarriers, and that the addition of ZnCO3 to the blocked-PLGA substantially slowed polymeric degradation. For example, in the rats injected with unblocked-PLGA microcarriers containing 0~ hGH and 0~ or 1 ZnCO3, no microcarriers were visible on day 21. In addition, for rats injected with blocked-P~GA microcarriers containing 0~ hGH and 0~ ZnCO3, a few microcarriers were visible on day 60 and none were visible on day so.
Furthermore, for rats injected with blocked-PLGA
microcarriers containing 0~ or 15~ hGH and 6~ ZnCO3, microcarriers were visible on day 90.

Exam~le 14 In Vivo Release of Aaqreqation-Stabilized hGH Microcarriers in Rats Studies were conducted in rats to screen various hGH
microcarrier formulations, determine pharmacokinetic parameters following intravenous (IV), subcutaneous (SC) and SC osmotic pump (Alzet~) administration of hGH, and to evaluate serum profiles and in vivo release rate of various hGH microcarrier formulations.
Sprague-Dawley rats were divided into groups of three each, randomized by body weight, and one hGH microcarrier formulation was administered to each group. Rats were CA 02223~83 1997-12-04 W O 96/40074 PCTrUS96/08526 injected subcutaneously with approximately 7.5 mg of hGH in 50 mg of microcarriers, suspended in 0.75 ml of an aqueous injection vehicle. The vehicle composition was 3~ CMC (low viscosity), 1~ Polysorbate 20, in 0.9~ NaCl. The microcarrier dose delivered was determined indirectly by weighing the residual dose in the injection vial and correcting for residual injection vehicle. The hGH dose was then computed from the protein loading of the microcarriers determined by nitrogen analysis.
Blood samples were collected at pre-determined intervals for up to 10 days after injection. Blood samples of 250 ~l were collected during the first 24 hours and at least 400 ~l at time points after 24 hours. Blood samples were clotted and hGH concentrations in serum were determined using a radio-immuno assay (RIA) using an RIA
kit from ICN.
For the determination of pharmacokinetic parameters, hGH in saline was administered to rats by subcutaneous bolus injection, intravenously or delivered via an osmotic pump which was implanted subcutaneously.
Three groups of rats received single subcutaneous injections of hGH in 0.9~ NaC1 at 0.5 or 7.5 mg/kg at a dose volume o~ 1.0 ml/kg and two groups received single intravenous bolus injections o~ hGH in 0.9~ NaC1 solution at about 1.0 mg and 5.0 mg of hGH per kg rat with a dose volume of 1.0 ml/kg. For the Alzet~ pump study, rats were divided into four groups of three rats each, randomized by body weight and dosed with about 20 mg/ml and 40 mg/ml hGH
in 0.9~ saline solution loaded into pumps (Alzet~ Model 2002, 200 ~l, 14 days release), and with about 4 mg/ml and 12 mg/ml hGH in 0.9~ saline solution loaded into pumps (Alzet Model 2ML4, 2ml, 28 days release). Expected release rates from the pumps correspond to about 2~ and 4 to 6~ of the ProLease hGH dose (about 15 mg/kg) per day, respectively. The Alzet pumps were implanted CA 02223~83 1997-12-04 subcutaneously in the inter-scapular region a~ter soaking for 1-2 minutes in sterile saline.
The ~ormulations o~ hGH sustained release microcarriers, synthesized as described in Example 5 contained 15~ w/w hGH complexed with Zn in a ratio o~ 6:1 Zn:hGH; 0~ , 3~ or 6~ w/w zinc carbonatei and 8K
unblocked PLGA, lOK blocked PLGA or 31K unblocked PLGA.
To evaluate the various hGH sustained release ~ormulations, Cmax, Cd5 and Cmax/Cd5 were the in vivo indices used, where Cmax is the maximum serum concentration observed, and Cd5 is the serum concentration at day 5 which should approximate the steady state concentration. The results were as ~ollows:

CA 02223~83 1997-12-04 W O 96/40074 PCT/U~5G/~~'~6 'Burst' Formulation in vi tro Monomer Cmax C day 5 Cmas/C

unblocked22.0 i 99 3*323.3 i 20.4 i19.5 i ~~ ZnC03 0.9 98.6 14.2 10.6 unblocked16.4 i 97 3*309-0 i 20 4 i39.5 i 1~ ZnC03 1.6 67.1 14.2 17.7 unblocked15.9 i 98 7 670.5 i 9.0 i44.8 i 3~ ZnC03 6.9 . 244.4 4.2 22.6 unblocked17.6 i 99 3 358.0 i 18.8 i42.4 i 6~ ZnC03 2.7 58.9 14.7 6.8 3lK PLGA
unblocked12.3 i 98 2 592 i 4.5 i132.5 i 0~ ZnC03 1.1 ' 318.2 1.5 47.9 3lK PLGA
unblocked11.4 i 98 8 432.7 i 5.1 i84.1 i 1~ ZnC03 1.3 91.6 0.3 14.9 unblocked7.9 i 1.999.4 643.6 i 8.0 i93.3 i 3~ ZnC03 203.9 2.6 62.0 3lK PLGA
unblocked15.8 i 99 8 1691.8 6.6 i262.2 i 2c 6~ ZnC03 0.5 340.0i 0.8 83.5 lOK PLGA
unblocked12.7 i 99 3 615.9 i 4.5 i155.0 i 1~ ZnC03 0.1 384.3 1.0 126.8 lOK PLGA
blocked 18.1 i 99 6 1053.2 3.6 ~291.7 i 3~ ZnC03 3.2 i 293.3 0.8 71.1 lOK PLGA
blocked9.9 i 1.4 99.0 1743.5 4.9 i516.1 i 6~ ZnC03 i 428.4 2.7 361.6 * Value obtained from duplicate batch of the same formulation.

The results of the screening showed that the two unblocked (8K and 31K) polymers had different in vivo CA 02223~83 1997-12-04 release kinetics compared to the original formulation, which used blocked lOK PLGA and 6~ w/w zinc carbonate.
Cmax values were generally lower with the unblocked polymer formulations than with the original formulation which suggested that the in vivo ~burst~ may be lower with the unblocked polymer formulations. The 'burst' was defined as the percent of hGH released in the first 24 hours after injection. The in vitro ~burst~ values were between 8-22~.
The zinc carbonate content of the formulations did not appear to have an effect on the ~burst' or the in vitro release profile.
The serum concentrations between days 4 and 6 were maintained at a fairly constant level above baseline (or the pre-bleed levels) with the unblocked polymer formulations, while serum concentrations with the blocked formulations, at the same time points were close to the baseline levels. The in vitro release data for up to 7 days showed that the released hGH protein was monomeric.
Useful data could not be obtained beyond day 6 because of anti-hGH antibody formation in the rats.

ExamPle 15 In Vivo Release of hGH from Aqqreqation-Stabilized hGH Microcarriers in ImmunosuP~ressed Rats Two groups of male Sprague-Dawley rats (N=3) (control groups), weighing 400+50g (S.D.) were injected as described in Example 14 with the microcarriers of Example 5. Two additional groups (N=3) of rats (test groups) were also given daily intraperitoneal injections of 10 mg cyclosporin A and 5 mg hydrocortisone in 0.5 ml sterilized saline for injection (USP) per kg of body weight for days 0 to 14 and then injections three times a week for days 15-28. No antibody titers were detected in these rats for the duration of treatment.

CA 02223~83 l997-l2-04 The control group did not receive injections to suppress their immune response to hGH. Antibodies were detected after day 6 in these rats.
The serum levels of hGH in the rats of the experimental groups and the control groups were determined by RIA through day 28. These results are provided in Figures 13 and 14. The results for both pairs of control and experimental groups were the same through day 6 suggesting that the cyclosporin A/hydrocortisone treatment did not affect the measured serum concentrations of hGH.
The results further show that the control groups' serum levels of hGH were artificially high due to their production of antibodies to hGH.
The results for the experimental groups, in which antibody formation was suppressed, showed sustained release of hGH for up to 24 days and 26 days for the 31K unblocked PLGA and 8K blocked PLGA microcarriers, respectively, of Example 5.

Exam~le 16 In Vivo Release of hGH From Aqqreaation-Stabilized hGH Microcarriers in Rhesus Monkeys The objective of this primate study was to evaluate the pharmacokinetic profiles of different hGH sustained release formulations as compared to more traditional methods of administering hGH (e.g., bolus sc injections, daily sc injections and sc injection combined with the use of an osmotic pump) and to determine which hGH sustained release formulation gave the optimal hGH blood concentration profile.
The formulations for the hGH sustained release microcarriers tested were 1) 15~ hGH (complexed with zn+2 at a 6:1 Zn+2:hGH ratio), 6~ w/w zinc carbonate and lOK
~locked PLGA; 2) 15~ hGH (complexed with zn+2 at a 6:1 Zn+2:hGH ratio), 1~ w/w zinc carbonate and 8K unblocked CA 02223~83 1997-12-04 PLGA ("RG502H" PLGA polymer); and 3) 15~ hGH (complexed with zn+2 at a 6:1 Zn+2:hGH ratio), 1~ w/w zinc carbonate and 31K unblocked PLGA ("RG503H" PLGA polymer). The microcarriers were formed as described in Example 5.
There were four monkeys per group and each ~n;m~l received a single subcutaneous injection into the dorsal cervical region on Day 1. A dose of 160 mg of hGH
sustained release microcarriers (24 mg of hGH) was administered to each monkey in 1.2 ml of injection vehicle through a 20 gauge needle. The injection vehicle was an aqueous vehicle cont~;n;ng 3~ w/v low viscosity Carboxymethyl Cellulose (sodium salt), 1~ v/v Tween 20 (Polysorbate 20) and 0.9~ sodium chloride.
The hGH dose was intended to provide measurable hGH
serum concentrations for pharmacokinetic analysis. To obtain pharmacokinetic parameters, additional study groups of four monkeys each were included, specifically 1) a single subcutaneous injection (24 mg hGH), 2) daily subcutaneous injections (24 mg/28 days = 0.86 mg hGH/day), 3) a subcutaneous injection (3.6 mg hGH) combined with an Alzet osmotic pump (20.4 mg hGH)(total dose of 24 mg hGH), and 4) a subcutaneous injection of the injection vehicle as a control (only used 3 monkeys for the vehicle control group).
The osmotic pump gave sustained serum hGH levels comparable to the hGH microcarriers up to day 28 as programmed to release hGH. The pumps were removed on day 31.
Blood samples were collected at the following times for hGH and IGF-l analyses: -7, -5, -3 days, pre-dose and, 0.5, 1, 2, 3, 5, 8, 10, 12, 24, 28, 32 and 48 hours, 5, 4, 6, 8, 11, 14, 17, 20, 23, 26, 29, 32, 25, 28, 41, 44, 47, 50, 53, 56 days post-dose.
The concentrations of IGF-l, which is expressed when a body has an effective serum level of hGH, and hGH in the CA 02223~83 1997-12-04 W O 96/40074 PCT~U59~3~6 serum were then measured. An IRMA kit from RADIM
(distributed by: Wein Laboratories, P.O. Box 227, Succasunna, NJ) was used to quantify hGH in monkey serum.
The IRMA assay had a limit of quantification in PBS buffer of 0.1 ng/mL and in pooled juvenile rhesus monkey serum of 1.5 ng/mL with a basal GH level of about 4ng/mL. RIA was used to quantify the IGF-1 serum levels.
The results of the hGH serum level assays for the lOK
blocked PLGA, 8K unblocked PLGA and 3lK unblocked hGH
microcarriers of Example 5 are provided in Figures 15-17, respectively. Further, the results of the hGH and IGF-1 serum assays for the 8K unblocked PLGA microcarriers of Example 5 are shown in Figure 18.
In addition, a comparison of the results of the IGF-1 serum assays for the 8K unblocked PLGA microcarriers of Example 5 as compared to the serum levels for daily subcutaneous injections of hGH are shown in Figure 19.
The results showed that the hGH sustained release microcarriers were releasing significant, sustained levels of hGH over a one month period while the subcutaneous injections were not able to maintain the same serum levels.
The IGF-1 serum profile showed that serum IGF-1 concentrations were elevated above the baseline values between days 2 and 29 after administering the microparticles. This shows that enough hGH was being released from the hGH sustained release microcarriers to cause a pharmacodynamic effect. This also indicates that the hGH released was biologically active which suggests that the encapsulation process had not adversely affected the biopotency of hGH.

CA 02223~83 1997-12-04 Example 17 In Vivo Release of Aqqreqation-Stabilized EPO from Polymeric Microcarriers in ImmunosuPpressed Rats Male Sprague-Dawley rats, weighing 400 + 50 g (S.D.), were used as the ~n; m~l model. The rats were not fasted before the experiments and subsequently were fed with a standard diet, an iron supplement, and allowed free access to water. Iron dextran (Sigma Co., St. Louis, MO) 5 mg/kg was injected intraperitoneally twice a week.
These experiments utilized the ;mmllnosuppression method described in Examples 11 and 15 for suppressing antibody production in the test ~n;m~l S in response to the EPO released (or injected) to obtain accurate profiles of serum EPO levels.
The purpose of the first experiment was to compare the in vivo pharmacodynamic effects of aggregation-stabilized EPO released from sustained release microcarriers to EPO
injected subcutaneously as a bolus, specifically upon serum reticulocyte profiles. Two groups of three rats were injected subcutaneously in the interscapular region on day O with 10,000 units of RMAm7 EPO microcarriers (unblocked lOK PLGA containing 10~ MgCO3 and 5~ Am7) and subsequently on day 28 with a 2,000 unit bolus of aqueous EPO. The control group did not receive the cyclosporin A/
hydrocortisone therapy, which the test group did receive.
Blood samples were taken from the tail vein of each rat at 1, 3, 4, 8, 10, 14, 16, 20, 24, 28, 30 or 31, 32 and 36 hours after injection. Additional blood samples were then taken approximately twice a week ~or the ~ollowing 5 weeks.
Blood reticulocyte levels were counted for selected blood sample. The results are provided in Figure 20.
Figure 20 shows higher reticulocyte counts in ;mmllnosuppressed rats in response to both the aggregation-stabilized EPO microcarriers and the EPO bolus. The non-CA 02223~83 1997-12-04 immunosuppressed rats (control group) showed lower reticulocyte levels due to antibody formation resulting from the immune systems' responses to EP0. This is particularly shown by the lack of a significant increase in reticulocyte levels in the control group after receiving the EP0 bolus on day 28.
Figure 20 also shows that injection with sustained release microcarriers resulted in a longer period of elevated serum reticulocyte levels than did a bolus of EP0.
lC The purpose of the second experiment was to compare the in vivo pharmacokinetic and pharmacodynamic effects of EP0 released from various sustained release microcarriers.
The rats in each of four groups rats (N=3) were injected subcutaneously in the interscapular region with one of four of the following formulations of microcarriers:

RMAml Unblocked lOK PLGA/10~ MgC03/5~ Aml RMMal Unblocked lOK PLGA 10~ MgC03/5~ Mal PZZnl Blocked lOK PLGA/10~ ZnC03/5~ Znl RMAm7 Unblocked lOK PLGA/10% MgC03/5~ Am7 Each rat received between 10,000 to 12,000 units per ~n;m~l . Each rat was also given daily an intraperitoneal injection of 10 mg of cyclosporin A and 5 mg of hydrocortisone.
Blood samples were taken from the tail vein of each rat at 1, 2, 4, 8, 10 (optionally), 24, 36 and 48 hours after injection. Additional blood samples were then taken approximately once a day for the following lO days and approximately two times per week for the next two weeks.
The EP0 concentration in the rat serum samples was determined using by ELISA. In addition, blood reticulocyte levels were counted.
Serum EP0 and blood reticulocyte profiles for these formulations are provided in Figures 21 and 22. EP0 levels CA 02223~83 1997-12-04 WO 96/40074 PCT/U',~ 6 r~m~;ned above baseline in these ~n;m~1s for approximately 14 days, showing a sustained release of biologically active EPO. Elevated reticulocyte levels were observed for about 17 days. Further, the response of immature and total reticulocyte levels were proportional and not significantly different from each other following EPO treatment.

Exam~le 18 Effect of Zinc Carbonate on Release Levels of Aaqreqation-Stabilized IFN-~,2b in Rats Rats (N=4) in three test groups were injected, as described in Example 9, with the microcarriers of formulas 4 and 6-8 of Example 2. The dose of IFN for each rat was about 0.8 mg/kg.
The purpose of the test was to determine if the initial burst and sustained level of IFN-~,2b released in vivo can be varied by changing the weight ratio of zinc carbonate to IFN-~,2b in microcarriers.
The weight ratio of zinc carbonate to IFN in microcarriers tested for initial burst effects were 0:1, 1:1, 3:1 and 8:1. Blood samples were then taken from the tail vein of each rat at 1, 2, 4, 8, 12, 24, 32, 48, 72, 96, 120, 144 and 168 hours after injection. The IFN-~,2b concentrations in the rat serum samples were determined by IRMA. The tests found that the addition of zinc carbonate to the formulation reduces initial burst in vivo.
Specifically, initial bursts measured, as a percentage of the total IFN in the microcarriers which were released over the first 24 hours, for microcarriers having weight ratios of 0:1, 1:1, 3:1 and 8:1 were 35+13~, 23+7~, 13+5~ and 8+1~, respectively.
These initial burst results suggest that the amount of metal cation in the polymer can be used to vary the burst.
For the sustained release test, the weight ratio of zinc carbonate to IFN in microcarriers tested were 1:1, 3:1 CA 02223~83 1997-12-04 W O 96/40074 PCT~US96/08526 and 8:1. The sustained release results of this test are presented in Figure 23. The sustained level observed for Formula 7 of Example 1, having a weight ratio of 1:1, was 250 + 30 IU/ml during days 5-7. The level observed for Formula 6, having a weight ratio of 3:1, was 180 + 10 IU/ml during days 5-7, whereas that for a Formula 8, having a weight ratio of 8:1, was 110 + 10 IU/ml.

EOUIVALENTS
Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to specific embodiments of the invention described specifically herein. Such equivalents are intended to be encompassed in the scope of the following claims.

Claims (27)

What is claimed is:

1. A device for the sustained release in vivo of a water soluble, biologically active agent wherein said agent is susceptible to aggregation, comprising.
a) a drug delivery device; and b) aggregation-stabilized, biologically active agent wherein said aggregation-stabilized agent is disposed within the drug delivery device.

2. A sustained release device of Claim 1 wherein the drug delivery device is a biocompatible polymeric matrix.

3. A sustained release device of Claim 1 wherein the aggregation-stabilized, biologically active agent includes a biologically active agent and an aggregation-stabilizer.

4. A sustained release device of Claim 3 wherein the aggregation-stabilizer is at least one salting-out salt.

5. A sustained release device of Claim 3 wherein the aggregation-stabilizer is a metal cation component.

6. A sustained release device of Claim 5 wherein the aggregation-stabilizer and a metal cation of the metal cation component are complexed.

7. A sustained release device of Claim 3 wherein the aggregation-stabilizer is a buffer.

8. A sustained release device of Claim 3 wherein the aggregation-stabilizer is polyethylene glycol.

9. A sustained release device of Claim 3 wherein the aggregation-stabilized, biologically active agent is in particulate form.

10. A composition for the sustained release in vivo of a water soluble, biologically active agent wherein said agent is susceptible to aggregation, comprising:
a) a biocompatible polymer; and b) aggregation-stabilized, biologically active agent wherein said aggregation-stabilized agent is disposed within the polymer.

11. A sustained release composition of Claim 10 wherein the aggregation-stabilized, biologically active agent includes a biologically active agent and an aggregation-stabilizer.

12. A sustained release composition of Claim 10 wherein the biologically active agent and the aggregation-stabilizer are mixed.

13. A sustained release composition of Claim 10 wherein the aggregation-stabilizer reduces the solubility of the protein in an aqueous fluid.

14. A composition of Claim 13 wherein the anti-aggregation agent is a salting-out salt.

15. A composition of Claim 14 wherein the salting-out salt comprises a salt containing a cation selected from the group consisting Mg+2, Li+, Na+, K+, NH4+ and combinations thereof.

16. A composition of Claim 13 wherein the salting-out salt comprises a salt containing an anion selected from the group consisting of SO4-2, HPO4-2, acetate, citrate, tartrate, CL-, NO3-, CLO3-, I-, CLO4-, SCN- and combinations thereof.

17. A composition of Claim 13 wherein the salting-out salt is ammonium sulfate.

18. A composition of Claim 11 wherein the aggregation stabilizer is mannitol.

19. A sustained release composition of Claim 11 wherein the aggregation-stabilizer is a buffer.

20. A sustained release composition of Claim 11 wherein the aggregation-stabilizer is a metal cation from a metal cation component.

21. A sustained release composition of Claim 11 wherein the metal cation of said metal cation component is a biocompatible multivalent cation.

22. A sustained release composition of Claim 12 wherein the multivalent metal cation is selected from the group consisting of Zn+2, Ca+2, Cu+2, Mg+2 and combinations thereof.

23. A sustained release composition of Claim 10 wherein the biocompatible polymeric matrix is formed of a biodegradable polymer.

24. A sustained release composition of Claim 20 further comprising a second metal cation component, wherein the second metal cation component is dispersed within the biocompatible polymeric matrix.

25. A sustained release composition of Claim 24 wherein the second metal cation component is selected from the group consisting of magnesium hydroxide, magnesium carbonate, calcium carbonate, zinc carbonate, magnesium acetate, zinc acetate, magnesium sulfate, zinc sulfate, magnesium chloride, zinc chloride, zinc citrate, magnesium citrate and a combination thereof.

26. A composition for the sustained release in vivo of a water soluble, biologically active agent aggregation, comprising:
a) a biodegradable polymer; and b) aggregation-stabilized, biologically active agent wherein said aggregation-stabilized agent is disposed within the biodegradable polymer.

27. A sustained release composition of Claim 26 wherein the biodegradable polymer is poly(lactide-co-glycolide).