Title:  Pulmonary administration of chemically modified insulin

United States Patent:  6,838,076

Issued:  January 4, 2005

Inventors:  Patton; John S. (Los Altos, CA); Kuo; Mei-Chang (Palo Alto, CA); Harris; J. Milton (Huntsville, AL); Leach; Chester (El Granada, CA); Perkins; Kimberly (Belmont, CA); Bueche; Blaine (Castro Valley, CA)

Assignee:  Nektar Therapeutics (San Carlos, CA)

Appl. No.:  154057

Filed:  May 21, 2002

Abstract

The present invention provides active, hydrophilic polymer-modified derivatives of insulin. The insulin derivatives of the invention are, in one aspect, suitable for delivery to the lung and exhibit pharmakokinetic and/or pharmacodynamic properties that are significantly improved over native insulin.

DETAILED DESCRIPTION OF THE INVENTION

The design, synthesis and characterization of various representative PEG-insulin conjugates have been optimized for pulmonary delivery to the lung. Although the preparation of PEG-insulin molecules has been previously described, the use of covalent coupling of PEG for providing prolonged action formulations of inhaleable insulin has not been previously demonstrated. In this regard, the challenge facing the applicants was to provide PEG-insulin conjugates having the optimal balance of number, location, structure, and molecular weight of PEG chains covalently attached to the insulin molecule to provide insulin compositions suitable for administration to the systemic circulation, preferably via the deep lung. Surprisingly, in light of the above, the inventors have discovered certain PEG- modified insulin formulations having one or more of the following features: (i) that are bioactive, i.e., that demonstrate at least about 5% of the activity of native insulin, or preferably have a bioactivity that is at least either substantially maintained or only minimally reduced from that of native insulin, or even more preferably, having an activity that is improved over native insulin, (ii) that are absorbed from the lung into the bloodstream (as opposed to "sticking" in the lung), (iii) that are chemically and physically stable, (iv) that, when administered to the lung, achieve blood levels of insulin that are elevated above baseline for at least about 8 hours post administration, (v) that are resistant to enzymatic attack by insulin-degrading enzymes, and (vi) that exhibit half lives that are extended over non-pegylated insulin when administered by inhalation, the details of which will become apparent when reading the following description.

Hydrophilic, Non-Naturally Occurring Polymer-Insulin Conjugates

Several illustrative PEG-insulin conjugates in accordance with the invention have been prepared. Although polyethylene glycol is a preferred polymer for use in the conjugates of the invention, other water-soluble, hydrophilic, non-naturally occurring polymers may also be employed. Other polymers suitable for use in the invention include polyvinylpyrrolidone, polyvinylalcohol, polyacryloylmorpholine, polyoxazoline, and poly(oxyethylated polyols) such as poly(oxyethylated glycerol), poly(oxyethylated sorbitol), and poly(oxyethylated glucose). Polymers comprising subunits or blocks of subunits selected from the above-described water-soluble polymers may also be used. Additionally, Co-polymers of polyethylene glycol and polypropylene glycol may be employed. Polymers of the invention are preferably substantially absent fatty acid groups or other lipophilic moieties.

The following section illustrates that with the careful selection of one or more PEG moieties, pegylation reagents, insulin pegylation sites, pegylation conditions and subsequent conjugate purification, PEG-insulin compositions with the desired clinical properties (improved pharmacokinetic and/or pharmacodynamic properties) can be obtained. Specific features of the PEG-insulin conjugates of the invention will now be provided.

A. Structural Features of the Polymer and the Resulting Conjugate

A PEG-insulin conjugate of the invention will typically comprise one or more PEG chains each having a molecular weight ranging from about 200 to about 40,000 daltons, and preferably ranging from about 200 to about 10,000 daltons. Preferably, a PEG for use in the invention will possess an average molecular weight falling within one of the following ranges: from about 200 to 10,000 daltons, from about 200 to about 7500 daltons, from about 200 to about 6,000 daltons, from about 200 to about 5,000 daltons, from about 200 to about 3000 daltons, from about 200 to about 2000 daltons, and from about 200 to about 1000 daltons.

Preferred PEG-insulins for administration to the lung will possess a PEG moiety having a molecular weight less than about 5000 daltons, preferably less than about 2000 daltons, and even less than about 1000 daltons. In one particular embodiment of the invention, the PEG-insulin conjugate possesses a PEG moiety having one of the following average molecular weights: 200, 300, 400, 500, 600, 750, 1000, 1500, 2000, 2500, 3000, 3500, 4000 or 5000. Higher molecular weight PEGs may, in certain instances, be less preferred due to a potential for loss of activity of the insulin molecule or, for pulmonary applications, reduced efficiency in crossing the lung.

While lower molecular weight PEGs may be preferred for increasing bioavailability, high molecular weight PEG chains, e.g., having an average molecular weight of 5,000, 10,000, 15,000, 20,000, 25,000, 30,000 or 40,000 daltons or greater, although generally found to decrease the bioavailability of native insulin, may be preferred for increasing half-life, particularly in the case of injectable formulations. That is to say, a significant improvement in the pharmacokinetic parameters, e.g., the area under the curve (AUC), for a high molecular weight PEG insulin (relative to native), can more than compensate for its diminished activity.

In terms of the number of subunits, PEGs for use in the invention will typically comprise a number of (OCH₂ CH₂) subunits falling within one or more of the following ranges: 2 to about 900 subunits, from about 4 to about 200 subunits, from about 4 to about 170 subunits, from about 4 to about 140 subunits, from about 4 to about 100 subunits, from about 10 to about 100 subunits, from about 4 to about 70 subunits, from about 4 to about 45 subunits, and from about 4 to about 25 subunits.

A PEG-insulin conjugate of the invention may be mono-substituted (i.e., that is to say, having a PEG attached to a single reactive insulin site) di-substituted (having PEG moieties attached to two reactive sites, tri-substituted, or even have polymer attachments at more than 3 sites on the insulin molecule. Mono-substituted, di-substituted, and tri-substituted insulin are also referred to herein as PEG monomer, dimer, and trimer, respectively. Multi-substituted insulin (meaning insulin having PEG moieties covalently attached at 2 or more insulin sites) will typically although not necessarily possess the same PEG moiety attached to each reactive site. That is to say, PEG-insulin compositions having more than one type of PEG moiety attached to insulin are contemplated. Preferred compositions in accordance with the invention are those containing predominantly monomer and/or dimer insulin conjugates. Surprisingly, PEG-insulin compositions that are not site-specific (comprising a mixture of PEG-insulin species having PEG covalently coupled to more than one reactive site) have been found to possess pharmacokinetic and pharmacodynamic properties improved over native insulin, in particular, when administered to the lung.

With respect to the position of PEG-substitution, the insulin molecule possesses several sites suitable for pegylation, with amino sites generally but not necessarily being most preferred. Specific insulin amino groups suitable for pegylation include the two N-termini, GlyA1 and PheB 1, as well as LysB29. These sites on the insulin molecule are also referred to herein simply as A1, B 1 and B29, respectively. Electrophilically activated PEGs for use in coupling to reactive amino groups on insulin include mPEG2-ALD, mPEG-succinimidyl propionate, mPEG-succinimidyl butanoate, mPEG-CM-HBA-NHA, mPEG-benzotriazole carbonate, mPEG-acetaldehyde diethyl acetal, and the like (Shearwater Corporation, Huntsville, Ala.).

A composition of the invention may, in one embodiment, contain predominantly (greater than 90%) monosubstituted insulin, e.g., mono-A1 insulin, mono-B1 insulin, or mono-B29 insulin. Such compositions may contain: i) mono-A1 insulin, ii) a mixture of mono-A1 insulin and mono-B1 insulin, or iii) a mixture of mono-A1, mono-B1 and mono-B29 insulin. Alternatively, a composition of the invention may contain predominantly di-substituted insulin, e.g., di-A1,B1-insulin, or di-A1,B29-insulin, or di-B1, B29-insulin, or any of the various combinations thereof.

Alternatively, a composition in accordance with the invention may contain a mixture of various PEG-insulin conjugates (i.e., PEG attached to any one of a combination of possible attachment sites). Using the amino sites on insulin as an example, a composition of the invention may contain any one or more of the following PEG-insulin conjugates: monoA1-PEG insulin, mono-B1-insulin, mono-B-29 insulin, di-A1, B1-insulin, di-A1,B29-insulin, di-B1,B29-insulin, or tri-A1,B-1,B29-insulin. In one embodiment, preferred are compositions containing predominantly monomers and dimers. Representative compositions may comprise PEG-insulin conjugates mixtures containing at least about 75% combined monomer and dimer, at least about 80% combined monomer and dimer, or at least about 85 to 90% combined monomer and dimer.

PheB 1 is a particularly preferred site for chemical modification by attachment of PEG. In particular, a PEG-insulin conjugate composition for use in the present invention may also be characterized in one embodiment as a composition in which at least about 70% of the B-1 sites on insulin are covalently coupled to PEG, regardless of the overall number of PEG-insulin species in the composition. Alternative embodiments include those in which at least about 75% of the B-1 sites on insulin are covalently coupled to PEG, or in which at least about 80% of the B-1 sites on insulin are covalently coupled to PEG, or in which at least about 90% or the B-1 sites on insulin are covalently coupled to PEG.

Surprisingly, the inventors have discovered that random mixtures of PEG-insulin (prepared by random rather than site-directed pegylation), when administered to the lung, result in elevated blood levels of insulin that are sustained for at least 6 hours, and more typically for at least 8 hours or greater post-administration. Such mixtures are advantageous not only due to their beneficial pharmacokinetic and pharmacodynamic properties, but because their synthesis is much simpler (does not require multiple synthetic steps, does not require the use of protecting groups, does not require multiple purifications, etc.) than the corresponding site-specific approach.

Alternative sites in the native insulin molecule that can be chemically modified by covalent attachment of PEG include the 2 C-termini, Arg22B, His10B, His5A, Glu4A, Glu17A, Glu13B, and Glu21B.

In addition to native insulin, non-native insulins having one or more amino acid substitutions, insertions, or deletions may be utilized such that additional sites become available for chemical modification by attachment of one or more PEG moieties. This embodiment of the invention is particularly useful for introducing additional, customized pegylation-sites within the insulin molecule, for example, for forming a PEG-insulin having improved resistance to enzymatic degradation. Such an approach provides greater flexibility in the design of an optimized insulin conjugate having the desired balance of activity, stability, solubility, and pharmacological properties. Although mutations can be carried out, i.e., by site specific mutagenesis, at any number of positions within the insulin molecule, preferred is an insulin variant in which any one of the first four amino acids in the B-chain is replaced with a cysteine residue. Such cysteine residues can then be reacted with an activated PEG that is specific for reaction with thiol groups, e.g., an N-maleimidyl polymer or other derivative, as described in U.S. Pat. No. 5,739,208 and in International Patent Publication No. WO 01/62827. Exemplary sulfhydryl-selective PEGs for use in this particular embodiment of the invention include mPEG-forked maleimide (mPEG(MAL)₂), mPEG2-forked maleimide (mPEG2(MAL)₂), mPEG-maleimide (mPEG-MAL), and mPEG2-maleimide (mPEG2-MAL) (Shearwater Corporation). The structures of these activated PEGS are as follows: mPEG-CONHCH [CH₂ CONH(CH₂ CH₂ O)₂ CH₂ CH₂ -MAL, mPEG2-lysine-NH--CH[CH₂ CONH (CH₂ CH₂ O)₂ CH₂ CH₂ -MAL]₂, mPEG-MAL, and mPEG2-lysine-NH--CH ₂ CH₂ NHC(O)CH₂ CH₂ MAL, respectively.

Additional mutations to the native insulin sequence may be employed, if necessary, to increase the bioactivity of a PEG-insulin conjugate whose biological activity is somewhat compromised as a result of pegylation. One such mutation is Thr8 to a His8. Additional mutations may be found, for example, in Diabetes Care, 13 (9), (1990), the content of which is herein incorporated by reference.

PEGs for use in the present invention may possess a variety of structures: linear, forked, branched, dumbbell, and the like. Typically, PEG is activated with a suitable activating group appropriate for coupling a desired site or sites on the insulin molecule. An activated PEG will possess a reactive group at a terminus for reaction with insulin. The term "linker" as used herein is meant to encompass an activating group positioned at a PEG terminus for reaction with insulin, and may further include additional (typically inert) atoms positioned between the PEG portion of the polymer and the activated group at the terminus, for ease in preparing the activated PEG. The linkers may contain any of a number of atoms, however, preferred are linkers containing methylenes intervening between the PEG backbone and the terminal activating group, e.g., as in mPEG-succinimidyl propionate and mPEG-butanoate. Representative activated PEG derivatives and methods for conjugating these agents to a drug such as insulin are known in the art and further described in Zalipsky, S., et al., "Use of Functionalized Poly(Ethylene Glycols) for Modification of Polypeptides" in Polyethylene Glycol Chemistry: Biotechnical and Biomedical Applications, J. M. Harris, Plenus Press, New York (1992), and in Advanced Drug Reviews, 16:157-182 (1995).

In one particular embodiment of the invention, the PEG portion of the conjugate is absent one or more lipophilic groups effective to significantly modify the water-soluble nature of the polymer or of the polymer-insulin conjugate. That is to say, the polymer or non-insulin portion of a conjugate of the invention may contain a group of atoms considered to be more lipophilic than hydrophilic (e.g., a carbon chain having from about 2 to 8-12 carbon atoms), however, if the presence of such a group or groups is not effective to significantly alter the hydrophilic nature of the polymer or of the conjugate, then such a moiety may be contained in the conjugates of the invention. That is to say, an insulin conjugate of the invention itself is characterized as hydrophilic, rather than lipophilic or amphiphilic. Typically, the polymer portion of an insulin conjugate, prior to coupling to insulin, whether or not containing such a lipid-loving group, will possess a high hydrophilic/lipophilic balance (HLB) number. The HLB number is based upon a weight percentage of each type of group (hydrophilic or lipophilic) in a molecule; values typically range from about 1-40. A polymer for use in the conjugates of the invention is, on a whole, characterized as hydrophilic, regardless of the presence of one or more lipid-loving substituents. In one embodiment of the invention, the polymer portion of a polymer-insulin conjugate is characterized by an HLB number of greater than 25 and more preferably greater than 30, or even more preferably greater than 35. In certain embodiments of the invention where such a lipophilic moiety may be present, the moiety is preferably not positioned at a terminus of a PEG chain.

Branched PEGs for use in the conjugates of the invention include those described in International Patent Publication WO 96/21469, the contents of which is expressly incorporated herein by reference in its entirety. Generally, branched PEGs can be represented by the formula R(PEG-OH)_n, where R represents the central "core" molecule and _n represents the number of arms. Branched PEGs have a central core from which extend 2 or more "PEG" arms. In a branched configuration, the branched polymer core possesses a single reactive site for attachment to insulin. Branched PEGs for use in the present invention will typically comprise fewer than 4 PEG arms, and more preferably, will comprise fewer than 3 PEG arms. Branched PEGs offer the advantage of having a single reactive site, coupled with a larger, more dense polymer cloud than their linear PEG counterparts. One particular type of branched PEG can be represented as (MeO-PEG-)_p R--X, where p equals 2 or 3, R is a central core structure such as lysine or glycerol having 2 or 3 PEG arms attached thereto, and X represents any suitable functional group that is or that can be activated for coupling to insulin. One particularly preferred branched PEG is mPEG2-NHS (Shearwater Corporation, Alabama) having the structure mPEG2-lysine-succinimide.

In yet another branched architecture, "pendant PEG" has reactive groups for protein coupling positioned along the PEG backbone rather than at the end of PEG chains as in the previous example. The reactive groups extending from the PEG backbone for coupling to insulin may be the same or different. Pendant PEG structures may be useful but are generally less preferred, particularly for compositions for inhalation.

Alternatively, the PEG-portion of a PEG-insulin conjugate may possess a forked structure having a branched moiety at one end of the polymer chain and two free reactive groups (or any multiple of 2) linked to the branched moiety for attachment to insulin. Exemplary forked PEGs are described in International Patent Publication No. WO 99/45964, the content of which is expressly incorporated herein by reference. The forked polyethylene glycol may optionally include an alkyl or "R" group at the opposing end of the polymer chain. More specifically, a forked PEG-insulin conjugate in accordance with the invention has the formula: R-PEG-L(Y-insulin)_n, where R is alkyl, L is a hydrolytically stable branch point and Y is a linking group that provides chemical linkage of the forked polymer to insulin, and n is a multiple of 2. L may represent a single "core" group, such as "--CH--", or may comprise a longer chain of atoms. Exemplary L groups include lysine, glycerol, pentaerythritol, or sorbitol. Typically, the particular branch atom within the branching moiety is carbon.

In one particular embodiment of the invention, the linkage of the forked PEG to the insulin molecule, (Y), is hydrolytically stable. In a preferred embodiment, n is 2. Suitable Y moieties, prior to conjugation with a reactive site on insulin, include but are not limited to active esters, active carbonates, aldehydes, isocyanates, isothiocyanates, epoxides, alcohols, maleimides, vinylsulfones, hydrazides, dithiopyridines, and iodacetamides. Selection of a suitable activating group will depend upon the intended site of attachment on the insulin molecule and can be readily determined by one of skill in the art. The corresponding Y group in the resulting PEG-insulin conjugate is that which results from reaction of the activated forked polymer with a suitable reactive site on insulin. The specific identity of such the final linkage will be apparent to one skilled in the art. For example, if the reactive forked PEG contains an activated ester, such as a succinimide or maleimide ester, conjugation via an amine site on insulin will result in formation of the corresponding amide linkage. These particular forked polymers are particularly attractive since they provide conjugates having a molar ratio of insulin to PEG of 2:1 or greater. Such conjugates may be less likely to block the insulin receptor site, while still providing the flexibility in design to protect the insulin against enzymatic degradation, e.g., by insulin degrading enzyme.

In a related embodiment, a forked PEG-insulin conjugate of the invention is represented by the formula: R-[PEG-L(Y-insulin)₂ ]_n. In this instance R represents a central core structure having attached thereto at least one PEG-di-insulin conjugate. Specifically, preferred forked polymers in accordance with this aspect of the invention are those were n is selected from the group consisting of 1,2,3,4,5,and 6. Exemplary core R structures may also be derived from lysine, glycerol, pentaerythritol, or sorbitol.

In an alternative embodiment, in any of the representative structures provided herein, the chemical linkage between insulin and the polymer branch point may be degradable (i.e., hydrolytically unstable). Alternatively, one or more degradable linkages may be contained in the polymer backbone to allow generation in vivo of a PEG-insulin conjugate having a smaller PEG chain than in the initially administered conjugate. Such optional features of the polymer conjugate may provide for additional control over the final desired pharmacological properties of the conjugate upon administration. For example, a large and relatively inert conjugate (i.e., having one or more high molecular weight PEG chains attached thereto, e.g., one or more PEG chains having a molecular weight greater than about 10,000, wherein the conjugate possesses essentially no bioactivity) may be administered, which then either in the lung or in the bloodstream, is hydrolyzed to generate a bioactive conjugate possessing a portion of the originally present PEG chain. In this way, the properties of the PEG-insulin conjugate may be somewhat more effectively tailored. For instance, absorption of the initial polymer conjugate may be slow upon initial administration, which is preferably but not necessarily by inhalation. Upon in-vivo cleavage of the hydrolytically degradable linkage, either free insulin (depending upon the position of the degradable linkage) or insulin having a small polyethylene tag attached thereto, is then released and more readily absorbed through the lung and/or circulated in the blood.

In one feature of this embodiment of the invention, the intact polymer-conjugate, prior to hydrolysis, is minimally degraded upon administration, such that hydrolysis of the cleavable bond is effective to govern the slow rate of release of active insulin into the bloodstream, as opposed to enzymatic degradation of insulin prior to its release into the systemic circulation.

Appropriate physiologically cleavable linkages include but are not limited to ester, carbonate ester, carbamate, sulfate, phosphate, acyloxyalkyl ether, acetal, and ketal. Such conjugates should possess a physiologically cleavable bond that is stable upon storage and upon administration. For instance, a PEG-cleavable linkage-insulin conjugate should maintain its integrity upon manufacturing of the final pharmaceutical composition, upon dissolution in an appropriate delivery vehicle, if employed, and upon administration irrespective of route.

More particularly, as described generally above, PEG-insulin conjugates having biodegradable linkages and useful in the present invention are represented by the following structures: PEG1-W-PEG2-insulin (where PEG1 and PEG2 can be the same or different) or PEG-W-insulin wherein W represents a weak, biodegradable linkage. These conjugates contain PEG arms or portions of PEG arms that are removable (i.e., cleavable) in-vivo. These particular modified insulins are typically substantially biologically inactive when intact, either due to the size of the intact PEG-portion of the molecule or due to steric blockage of the active sites on the insulin molecule by the PEG chain. However, such conjugates are cleaved under physiological conditions to thereby release insulin or a biologically active PEG-insulin capable of absorption into the systemic circulation, e.g., from the lung. In a first exemplary structure, the PEG1 portion may possess any of a number of different architectures discussed herein, and will typically possess a molecular weight of at least about 10,000, such that the conjugate is not rapidly absorbed upon administration. The PEG2 portion of the molecule preferably possesses a molecular weight of less than about 5000 daltons, more preferably less than 2000 daltons, and even more preferably less than 1000 daltons. Referring now to the secondary exemplary structure, PEG-W-insulin, the PEG portion will generally possess a molecular weight of at least about 10,000 Daltons or more.

In yet another specific embodiment of the invention, the PEG-insulin conjugate has a dumbbell-like structure in which two insulin moieties are interconnected by a central PEG. More specifically, such conjugates may be represented by the structure insulin-Y-PEG-Z-insulin, where Y and Z are hyrolytically stable linking groups linking insulin to the PEG moiety. In a particular embodiment, Z is an activated sulfone, which prior to conjugation, is suitable for reaction with thiol groups on insulin (e.g., cysteines). Alternatively, Y and Z may be any group suitable for covalent coupling with insulin. Additional examples are provided in U.S. Pat. No. 5,900,461, the content of which is expressly incorporated herein by reference.

Additional representative mono-and di-functional PEGs having either linear or branched structures for use in preparing the conjugates of the invention may be purchased from Shearwater Corporation (Alabama). Illustrative structures are described in Shearwater's 2001 catalogue entitled "Polyethylene Glycol and Derivatives for Biomedical Applications", the contents of which is expressly incorporated herein by reference.

B. Preparation

The reaction conditions for coupling PEG to insulin will vary depending upon the particular PEG derivative employed, the site of attachment on insulin and the particular type of reactive group (i.e., lysine versus cysteine), the desired degree of pegylation, and the like, and can readily be determined by one skilled in the art.

As exemplified in greater detail below, synthesis of the conjugates of the invention may be site-directed or may be random. Suitable PEG activating groups for reaction with insulin amine groups (e.g., GlyA1, PheB1, Lys29B), are tresylate, aldehyde, epoxide, carbonylimidazole, active carbonates (e.g. succinimidyl carbonate), acetal, and active esters such as N-hydroxylsuccinimide (NHS) and NHS-derivatized PEGs . Of these, the most reactive are PEG carboxymethyl-NHS, norleucine-NHS, and succinimidyl carbonate. Additional PEG reagents for coupling to insulin include PEG succinimidyl succinate and propionate. PEG active esters suitable for use in the invention, e.g., having a single propanoic or butanoic acid moiety, are described in U.S. Pat. No. 5,672,662, the contents of which is incorporated herein in its entirety. Specific active esters for use in preparing the conjugates of the invention include mPEG-succinimidyl propionate and mPEG-succinimidyl butanoate.

Optimized experimental conditions for a specific conjugate can readily be determined, typically by routine experimentation, by one skilled in the art.

Reactive groups suitable for activating a PEG-polymer for attachment to a thiol (sulfhydryl) group on insulin include vinylsulfones, iodoacetamide, maleimide, and dithio-orthopyridine. Particularly preferred reagents include PEG vinylsulfones and PEG-maleimide. Additional representative vinylsulfones for use in the present invention are described in U.S. Pat. No. 5,739,208, the content of which is expressly incorporated herein by reference.

In some instances, the compositions of the invention comprise selectively PEGylated insulin, i.e., the resulting conjugates are essentially homogeneous with respect to the position and degree of pegylation. That is to say, site selective or site directed pegylation of an amino group will result in an insulin conjugate composition wherein primarily the intended target position, e.g., PheB1, has a PEG moiety attached thereto. Depending upon the intended site of pegylation, a protection/deprotection synthetic strategy may be necessary to prevent pegylation of non-target reactive sites within the insulin molecule, e.g., by employing a protecting group such as t-BOC (tert-butoxycarbonyl) or di-BOC (di- butoxycarbonyl). Other suitable amino protecting groups include carbobenzoxy (CBZ), trityl derivatives such as trityl (Tr), dimethoxytrityl (DMTr) and the like. Other protecting groups, such as cyclic diacyl groups or nitrophenylsulfenyl (Nps) may also prove useful for protecting amino functions.

Such site directed coupling chemistry employed to provide the insulin conjugates of the invention results in compositions having a large degree of substitution at a particular reactive site on the insulin molecule. These compositions can then, if desired, be further purified to provide compositions of essentially pure mono- or di-functional PEG-insulins.

An essentially pure PEG-insulin composition refers to one comprising a PEG-insulin conjugate that is at least about 90% pure, and preferably at least about 95% pure by any one of the following analytical methods. In this respect, purity refers to PEG-insulin conjugate content. That is to say, a PEG-insulin conjugate that is at least about 90% pure contains at least about 90% by weight of PEG-insulin conjugate species, while the other nearly 10% represents impurities that are not PEG-insulin conjugate. PEG-insulin conjugates of the invention are typically purified using one or more purification techniques such as ion exchange chromatography, size exclusion chromatography, affinity chromatography, hydrophobic interaction chromatography, and reverse phase chromatography. The overall homogeneity of the resulting PEG-insulin (number of insulin-PEG forms present) can be assessed using one or more of the following methods: chromatography, electrophoresis, mass spectrometry, and in particular, MALDI-MS, and NMR spectroscopy. One particularly useful method for identifying the sites of insulin modification is RP-HPLC peptide mapping, coupled with a USP identity test for human insulin using endoproteinase Glu-C.

C. Characteristics of PEG-insulin Conjugates

In accordance with one aspect of the invention, provided are PEG-insulin conjugate compositions that are suitable for pulmonary administration.  The PEG insulin conjugates of the invention, when administered to the lung, possess pharmacokinetic and pharmodynamic properties improved over native insulin. It has been shown that insulin can be modified with PEGs having a molecular weight of up to 5,000K to 10,000 K or greater, and still maintain activity.   Additionally, as can be seen from the examples provided herein, exemplary PEG-insulin conjugates possessing PEG chains with average molecular weights ranging from 750 daltons, to 2,000 daltons, to 5,000 daltons, when administered both intravenously and to the lung, are bioactive, are not substantially held up within the lung when administered to the lung, as evidenced by detectable serum levels of insulin, and are effective in producing a substantial suppression of glucose, which, in certain cases, is over a duration of time significantly greater than that observed for native insulin. Moreover, provided herein are PEG-insulin conjugates, which when administered to the lung, exhibit a rapid onset of action (within 1 hour of administration).

In general, a PEG-insulin composition of the invention will possess one or more of the following characteristics. The PEG-insulin conjugates of the invention maintain at least a measurable degree of specific activity. That is to say, a PEG-insulin conjugate in accordance with the invention will possesses anywhere from about 2% to about 100% or more of the specific activity of native insulin. In one preferred embodiment of the invention, the PEG-insulin conjugate will possess at least 10% or more of the biological activity of unmodified, native insulin and is substantially non-immunogenic. Preferably, the bioactivity of a conjugate of the invention will range from about 5% to at least about 20% or more of the bioactivity of native insulin. The bioactivity of a conjugate of the invention may be characterized indirectly, e.g., by monitoring blood glucose and insulin levels to generate the corresponding pharmacodynamic and/or pharmacokinetic data, or by RIA (radioimmunoassay).

In considering serum concentrations of insulin following administration of a PEG-insulin conjugate, e.g., to the lung, the conjugates described herein will typically peak (i.e., reach Cmax or the highest point in the concentration curve) at from around 2 to 8 hours post dose, and more typically will peak at around 3 to 6 hours or so. Moreover, the chemically modified insulins of the invention, and in particular, the prolonged effect insulin formulations provided herein, are effective in providing both a measurable glucose-lowering effect and sustained concentrations of insulin over a longer period of time than native insulin. More specifically, a PEG-insulin conjugate when administered to the lung will exhibit elevated levels of insulin (elevated over basal or baseline levels) for at least about 6 hours and preferably for at least 8 hours post administration. Preferably, a PEG-insulin conjugate when administered to the lung, results in elevated blood levels of insulin over a prolonged period of at least 9 hours, 10 hours, 12 hours or at least 14 hours post administration wherein above-basal levels of insulin conjugate are detectable in the bloodstream for such an extended duration post dose.

As described previously, an insulin conjugate of the invention is effective to lower blood glucose levels. Turning now to the ability of the compositions of the invention to suppress blood glucose, a PEG insulin conjugate when administered, e.g., to the lung, is effective to suppress blood glucose levels below basal levels for at least 6 hours post-administration. More particularly, a PEG-insulin composition of the invention is effective to suppress blood glucose levels below baseline for at least 8 hours, preferably for at least 10 hours, or more preferably for at least 12 hours or more post administration.

Moreover, the PEG-insulin formulations of the invention exhibit absolute pulmonary bioavailabilities that are improved over native insulin. Specifically, a PEG-insulin formulation as provided herein possesses an absolute pulmonary bioavailability that is at least about 1.2 times that of native insulin, preferably at least about 1.5 times that of native insulin, more preferably is at least about 2 times greater or even more preferably is at least about 2.5 or 3 times greater than that of native insulin. (Illustrative results are provided in Table 13).

III. Formulations

The polymer-insulin conjugate compositions of the invention may be administered neat or in therapeutic/pharmaceutical compositions containing additional excipients, solvents, stabilizers, etc., depending upon the particular mode of admistration and dosage form. The present conjugates may be administered parenterally as well as non-parenterally. Specific administration routes include oral, rectal, buccal, topical, nasal, ophthalmic, subcutaneous, intramuscular, intraveneous, transdermal, and pulmonary. Most preferred are parenteral and pulmonary routes.

Pharmaceutical formulations for mammalian and preferably human administration will typically comprise at least one PEG-insulin conjugate of the invention together with one or more pharmaceutically acceptable carriers, as will be described in greater detail below, particularly for pulmonary compositions. Formulations of the present invention, e.g., for parenteral administration, are most typically liquid solutions or suspensions, while inhaleable formulations for pulmonary administration are generally liquids or powders, with powder formulations being generally preferred. Additional albeit less preferred compositions of the chemically modified insulins of the invention include syrups, creams, ointments, tablets, and the like.

Formulations and corresponding doses of insulin will vary with the concentration bioactivity of the insulin employed. Injectable insulin is measured in USP Insulin Units and USP Insulin Human Units (U); one unit of insulin is equal to the amount required to reduce the concentration of blood glucose in a fasting rabbit to 45 mg/dl (2.5 mM). Typical concentrations of insulin preparations for injection range from 30-100 Units/mL which is about 3.6 mg of insulin per mL. The amount of insulin required to achieve the desired physiological effect in a patient will vary not only with the particulars of the patient and his disease (e.g., type I vs. type II diabetes) but also with the strength and particular type of insulin used. For instance, dosage ranges for regular insulin (rapid acting) are from about 2 to 0.3 U insulin per kilogram of body weight per day. Compositions of the invention are, in one aspect, effective to achieve in patients undergoing therapy a fasting blood glucose concentration between about 90 and 140 mg/dl and a postprandial value below about 250 mg/dl. Precise dosages can be determined by one skilled in the art when coupled with the pharmacodynamics and pharmacokinetics of the precise insulin-conjugate employed for a particular route of administration, and can readily be adjusted in response to periodic glucose monitoring.

Individual dosages (on a per inhalation basis) for inhaleable insulin-conjugate formulations are typically in the range of from about 0.5 mg to 15 mg insulin-conjugate, where the desired overall dosage is typically achieved in from about 1-10 breaths, and preferably in from about 1 to 4 breaths. On average, the overall dose of PEG-insulin administered by inhalation per dosing session will range from about 10U to about 400U, with each individual dosage or unit dosage form (corresponding to a single inhalation) containing from about 5U to 400U.

A. Inhaleable Formulations of Chemically Modified Insulin

As stated above, one preferred route of administration for the insulin conjugates of the invention is by inhalation to the lung. Particular formulation components, characteristics and delivery devices will now be more fully described.

The amount of insulin conjugate in the formulation will be that amount necessary to deliver a therapeutically effective amount of insulin per unit dose to achieve at least one of the therapeutic effects of native insulin, i.e., the ability to control blood glucose levels to near normoglycemia. In practice, this will vary widely depending upon the particular insulin conjugate, its activity, the severity of the diabetic condition to be treated, the patient population, the stability of the formulation, and the like. The composition will generally contain anywhere from about 1% by weight to about 99% by weight PEG-insulin, typically from about 2% to about 95% by weight conjugate, and more typically from about 5% to 85% by weight conjugate, and will also depend upon the relative amounts of excipients/additives contained in the composition. More specifically, the composition will typically contain at least about one of the following percentages of PEG-insulin: 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or more by weight. Preferably, powder compositions will contain at least about 60%, e.g., from about 60-100% by weight PEG-insulin. It is to be understood that more than one insulin may be incorporated into the formulations described herein and that the use of the term "agent" or "insulin" in no way excludes the use of two or more insulins or a combination of insulin with another active agent. (For example, an illustrative PEG-insulin formulation may also comprise native insulin).

A.1. Excipients

Compositions of the invention will, in most instances, include one or more excipients. Preferred are carbohydrate excipients, either alone or in combination with other excipients or additives. Representative carbohydrates for use in the compositions of the invention include sugars, derivatized sugars such as alditols, aldonic acids, esterified sugars, and sugar polymers. Exemplary carbohydrate excipients suitable for use in the invention include, for example, monosaccharides such as fructose, maltose, galactose, glucose, D-mannose, sorbose, and the like; disaccharides, such as lactose, sucrose, trehalose, cellobiose, and the like; polysaccharides, such as raffinose, melezitose, maltodextrins, dextrans, starches, and the like; and alditols, such as mannitol, xylitol, maltitol, lactitol, xylitol sorbitol (glucitol), pyranosyl sorbitol, myoinositol and the like. Preferred are non-reducing sugars, sugars that can form a substantially dry amorphous or glassy phase when combined with an insulin conjugate, and sugars possessing relatively high Tgs (e.g., Tgs greater than 40^oC., preferably greater than 50^oC., more preferably greater than 60^oC., and even more preferably greater than 70^oC., and most preferably having Tgs of 80^oC. and above).

Additional excipients include amino acids, peptides and particularly oligomers comprising 2-9 amino acids, and more preferably 2-5 mers, and polypeptides, all of which may be homo or hetero species. Representative amino acids include glycine (gly), alanine (ala), valine (val), leucine (leu), isoleucine (ile), methionine (met), proline (pro), phenylalanine (phe), trytophan (trp), serine (ser), threonine (thr), cysteine (cys), tyrosine (tyr), asparagine (asp), glutamic acid (glu), lysine (lys), arginine (arg), histidine (his), norleucine (nor), and modified forms thereof. One particularly preferred amino acid is leucine.

Also preferred for use as excipients in inhaleable compositions are di- and tripeptides containing two or more leucyl residues, as described in Inhale Therapeutic System's International patent application PCT/US00/09785, incorporated herein by reference in its entirety.

Also preferred are di- and tripeptides having a glass transition temperature greater than about 40^oC., more preferably greater than 50^oC., even more preferably greater than 60^oC., and most preferably greater than 70^oC.

Although less preferred due to their limited solubility in water, additional stability and aerosol performance-enhancing peptides for use in the invention are 4-mers and 5-mers containing any combination of amino acids as described above. More preferably, the 4-mer or 5-mer will comprise two or more leucine residues. The leucine residues may occupy any position within the peptide, while the remaining (i.e., non-leucyl) amino acids positions are occupied by any amino acid as described above, provided that the resulting 4-mer or 5-mer has a solubility in water of at least about 1 mg/ml. Preferably, the non-leucyl amino acids in a 4-mer or 5-mer are hydrophilic amino acids such as lysine, to thereby increase the solubility of the peptide in water.

Polyamino acids, and in particular, those comprising any of the herein described amino acids, are also suitable for use as stabilizers. Preferred are polyamino acids such as poly-lysine, poly-glutamic acid, and poly(lys, ala).

Additional excipients and additives useful in the present compositions and methods include but are not limited to proteins, non-biological polymers, and biological polymers, which may be present singly or in combination. Suitable excipients are those provided in Inhale Therapeutic Systems' International Publication Nos. WO 96/32096 and 98/16205. Preferred are excipients having glass transition temperatures (Tg), above about 35^oC., preferably above about 40^oC., more preferably above 45^oC., most preferably above about 55^oC.

Exemplary protein excipients include albumins such as human serum albumin (HSA), recombinant human albumin (rHA), gelatin, casein, hemoglobin, and the like. The compositions may also include a buffer or a pH adjusting agent, typically but not necessarily a salt prepared from an organic acid or base. Representative buffers include organic acid salts of citric acid, ascorbic acid, gluconic acid, carbonic acid, tartaric acid, succinic acid, acetic acid, or phthalic acid. Other suitable buffers include Tris, tromethamine hydrochloride, borate, glycerol phosphate and phosphate. Amino acids such as glycine are also suitable.

The compositions of the invention may also include additional polymeric excipients/additives, e.g., polyvinylpyrrolidones, derivatized celluloses such as hydroxymethylcellulose, hydroxyethylcellulose, and hydroxypropylmethylcellulose, Ficolls (a polymeric sugar), hydroxyethylstarch (HES), dextrates (e.g., cyclodextrins, such as 2-hydroxypropyl-.beta.-cyclodextrin and sulfobutylether-.beta.-cyclodextrin), polyethylene glycols, and pectin.

The compositions may further include flavoring agents, taste-masking agents, inorganic salts (e.g., sodium chloride), antimicrobial agents (e.g., benzalkonium chloride), sweeteners, antioxidants, antistatic agents, surfactants (e.g., polysorbates such as "TWEEN 20" and "TWEEN 80", and pluronics such as F68 and F88, available from BASF), sorbitan esters, lipids (e.g., phospholipids such as lecithin and other phosphatidylcholines, phosphatidylethanolamines, although preferably not in liposomal form), fatty acids and fatty esters, steroids (e.g., cholesterol), and chelating agents (e.g., EDTA, zinc and other such suitable cations). The use of certain di-substituted phosphatidylcholines for producing perforated microstructures (i.e., hollow, porous microspheres) is described in greater detail below. Other pharmaceutical excipients and/or additives suitable for use in the compositions according to the invention are listed in "Remington: The Science & Practice of Pharmacy", 19^th ed., Williams & Williams, (1995), and in the "Physician's Desk Reference", 52^nd ed., Medical Economics, Montvale, N.J. (1998).

In one embodiment, a composition in accordance with the invention may be absent penetration enhancers, which can cause irritation and are toxic at the high levels often necessary to provide substantial enhancement of absorption. Specific enhancers, which may be absent from the compositions of the invention, are the detergent-like enhancers such as deoxycholate, laureth-9, DDPC, glycocholate, and the fusidates. Certain enhancers, however, such as those that protect insulin from enzyme degradation, e.g., protease and peptidase inhibitors such as alpha-i antiprotease, captropril, thiorphan, and the HIV protease inhibitors, may, in certain embodiments of the invention, be incorporated in the PEG-insulin formulations of the invention. In yet another embodiment, the PEG-insulin conjugates of the invention may be absent liposomes, lipid matrices, and encapsulating agents.

Generally, the pharmaceutical compositions of the invention will contain from about 1% to about 99% by weight excipient, preferably from about 5%-98% by weight excipient, more preferably from about 15-95% by weight excipient. Even more preferably, the spray dried composition will contain from about 0-50% by weight excipient, more preferably from 0-40% by weight excipient. In general, a high insulin concentration is desired in the final pharmaceutical composition. Typically, the optimal amount of excipient/additive is determined experimentally, i.e., by preparing compositions containing varying amounts of excipients (ranging from low to high), examining the chemical and physical stability of the PEG-insulin, MMADs and dispersibilities of the pharmaceutical compositions, and then further exploring the range at which optimal aerosol performance is attained with no significant adverse effect upon insulin stability.

A.2. Preparing Dry Powders

Dry powder formulations of the invention comprising a PEG-insulin conjugate may be prepared by any of a number of drying techniques, and preferably by spray drying. Spray drying of the formulations is carried out, for example, as described generally in the "Spray Drying Handbook", 5^th ed., K. Masters, John Wiley & Sons, Inc., NY, N.Y. (1991), and in Platz, R., et al., International Patent Publication Nos. WO 97/41833 (1997) and WO 96/32149 (1996), the contents of which are incorporated herein by reference.

Solutions of PEG-insulin conjugates are spray dried in a conventional spray drier, such as those available from commercial suppliers such as Niro A/S (Denmark), Buchi (Switzerland) and the like, resulting in a dispersible, dry powder. Optimal conditions for spray drying the PEG-insulin solutions will vary depending upon the formulation components, and are generally determined experimentally. The gas used to spray dry the material is typically air, although inert gases such as nitrogen or argon are also suitable. Moreover, the temperature of both the inlet and outlet of the gas used to dry the sprayed material is such that it does not cause degradation of the PEG-insulin in the sprayed material. Such temperatures are typically determined experimentally, although generally, the inlet temperature will range from about 50^oC. to about 200^oC., while the outlet temperature will range from about 30^oC. to about 150^oC. Preferred parameters include atomization pressures ranging from about 20-150 psi, and preferably from about 30-40 to 100 psi. Typically the atomization pressure employed will be one of the following (psi): 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120 or above.

Respirable PEG-insulin compositions having the features described herein may also be produced by drying certain formulation components which result in formation of a perforated microstructure powder as described in WO 99/16419, the entire contents of which are incorporated by reference herein. The perforated microstructure powders typically comprise spray-dried, hollow microspheres having a relatively thin porous wall defining a large internal void. The perforated microstructure powders may be dispersed in a selected suspension media (such as a non-aqueous and/or fluorinated blowing agent) to provide stabilized dispersions prior to drying. The use of relatively low density perforated (or porous) microstructures or microparticulates significantly reduces attractive forces between the particles, thereby lowering the shear forces, increasing the flowability and dispersibility of the resulting powders, and reducing the degradation by flocculation, sedimentation or creaming of the stabilized dispersions thereof.

Alternatively, a PEG-insulin composition for pulmonary delivery may comprise aerodynamically light particles as described in U.S. Pat. No. 6,136,295.

A powdered formulation of the invention may also be prepared by lyophilization, vacuum drying, spray freeze drying, super critical fluid processing (e.g., as described in Hanna, et al., U.S. Pat. No. 6,063,138), air drying, or other forms of evaporative drying.

In yet another approach, dry powders may be prepared by agglomerating the powder components, sieving the materials to obtain agglomerates, spheronizing to provide a more spherical agglomerate, and sizing to obtain a uniformly-sized product, as described, e.g., and in Ahlneck, C., et al., International PCT Publication No. WO 95/09616, 1995, incorporated herein by reference.

Dry powders may also be prepared by blending, grinding, sieving or jet milling formulation components in dry powder form.

Once formed, the dry powder compositions are preferably maintained under dry (i.e., relatively low humidity) conditions during manufacture, processing, and storage. Irrespective of the drying process employed, the process will preferably result in inhaleable, highly dispersible particles comprising a chemically modified insulin as described herein.

A.3. Features of Dry Powder Formulations

Powders of the invention are further characterized by several features, most notably, one or more of the following: (i) consistently high dispersibilities, which are maintained, even upon storage (ii) small aerodynamic particles sizes (MMADs), (iii) improved fine particle dose values, i.e., powders having a higher percentage of particles sized less than 3.3 microns MMAD, all of which contribute to the improved ability of the powder to penetrate to the tissues of the lower respiratory tract (i.e., the alveoli) for delivery to the systemic circulation. These physical characteristics of the inhaleable powders of the invention, to be described more fully below, are important in maximizing the efficiency of aerosolized delivery of such powders to the deep lung.

Dry powders of the invention are composed of aerosolizable particles effective to penetrate into the lungs. The particles of the invention have a mass median diameter (MMD) of less than about 20-30 .mu.m, or less than 20 .mu.m, or less than about 10 .mu.m, preferably less than about 7.5 .mu.m, and more preferably less than about 4 .mu.m, and even less than about 3.5 .mu.m, and usually are in the range of 0.1 .mu.m to 5 .mu.m in diameter. Preferred powders are composed of particles having an MMD from about 0.2 to 4.0 .mu.m. In some cases, the powder will also contain non-respirable carrier particles such as lactose, where the non-respirable particles are typically greater than about 40 microns in size.

The powders of the invention are further characterized by an aerosol particle size distribution less than about 10 .mu.m mass median aerodynamic diameter (MMAD), preferably having MMADs less than about 5 .mu.m, more preferably less than 4.0 .mu.m, even more preferably less than 3.5 .mu.m, and most preferably less than 3 .mu.m. The mass median aerodynamic diameters of the powders will characteristically range from about 0.1-10 .mu.m, preferably from about 0.2-5.0 .mu.m MMAD, more preferably from about 1.0-4.0 .mu.m MMAD, and even more preferably from about 1.5 to 3.0 .mu.m. Small aerodynamic diameters can generally be achieved by a combination of optimized spray drying conditions and choice and concentration of excipients.

The PEG-insulin powders of the invention may further be characterized by their densities. A powdered composition for inhalation will generally possess a bulk density from about 0.1 to 10 g/cubic centimeter, preferably from about 0.1-2 g/cubic centimeter, and more preferably from about 0.15-1.5 g/cubic centimeter.

The powders will generally have a moisture content below about 20% by weight, usually below about 10% by weight, and preferably below about 5% by weight. Preferred powders in accordance with the invention having a moisture content that is below about one or more of the following weight percentages: 15%, 10%, 7%, 5%, or 3%. Such low moisture-containing solids tend to exhibit a greater stability upon packaging and storage.

Additionally, the spray drying methods and stabilizers described herein are effective to provide highly dispersible PEG-insulin formulations. For powder formulations, the emitted dose (ED) of these powders is typically greater than 30%, and usually greater than 40%. More preferably, the ED of the powders of the invention is greater than 50%, and is often greater than 60%.

The compositions described herein also possess good stability with respect to both chemical stability and physical stability, i.e., aerosol performance over time. Generally, with respect to chemical stability, the PEG-insulin conjugate contained in the formulation will degrade by no more than about 15% upon spray drying. That is to say, the powder will possess at least about 85% intact PEG-insulin conjugate, preferably at least about 90 or 95% intact conjugate, and even more preferably will contain at least about 97% or greater intact PEG-insulin. Preferably, the spray drying process will result in powders having less than about 10% total protein aggregates, that is to say, greater than 90% by weight of the chemically modified insulin being in monomeric form.

With respect to aerosol performance, compositions of the invention are generally characterized by a drop in emitted dose of no more than about 20%, preferably no more than about 15%, and more preferably by no more than about 10%, when stored under ambient conditions for a period of three months.

A.4. Administration of the Composition

PEG-insulin formulations as described herein may be delivered using any suitable dry powder inhaler (DPI), i.e., an inhaler device that utilizes the patient's inhaled breath as a vehicle to transport the dry powder drug to the lungs. Preferred are Inhale Therapeutic Systems' dry powder inhalation devices as described in Patton, J. S., et al., U.S. Pat. No. 5,458,135, Oct. 17, 1995; Smith, A. E., et al., U.S. Pat. No. 5,740,794, Apr. 21, 1998; and in Smith, A. E., et. al., U.S. Pat. No. 5,785,049, Jul. 28, 1998, herein incorporated by reference. When administered using a device of this type, the powdered medicament is contained in a receptacle having a puncturable lid or other access surface, preferably a blister package or cartridge, where the receptacle may contain a single dosage unit or multiple dosage units. Convenient methods for filling large numbers of cavities (i.e., unit dose packages) with metered doses of dry powder medicament are described, e.g., in Parks, D. J., et al., International Patent Publication WO 97/41031, Nov. 6, 1997, incorporated herein by reference.

Other dry powder dispersion devices for pulmonary administration of dry powders include those described, for example, in Newell, R. E., et al, European Patent No. EP 129985, Sep. 7, 1988; in Hodson, P. D., et al., European Patent No. EP472598, Jul. 3, 1996; in Cocozza, S., et al., European Patent No. EP 467172, Apr. 6, 1994, and in Lloyd, L. J. et al., U.S. Pat. No. 5,522,385, Jun. 4, 1996, incorporated herein by reference. Also suitable for delivering PEG-insulin dry powders are inhalation devices such as the Astra-Draco "TURBUHALER". This type of device is described in detail in Virtanen, R., U.S. Pat. No. 4,668,218, May 26, 1987; in Wetterlin, K., et al., U.S. Pat. No. 4,667,668, May 26, 1987; and in Wetterlin, K., et al., U.S. Pat. No. 4,805,811, Feb. 21, 1989, all of which are incorporated herein by reference. Other suitable devices include dry powder inhalers such as Rotahaler.RTM. (Glaxo), Discus.RTM. (Glaxo), Spiros.TM. inhaler (Dura Pharmaceuticals), and the Spinhaler.RTM. (Fisons). Also suitable are devices which employ the use of a piston to provide air for either entraining powdered medicament, lifting medicament from a carrier screen by passing air through the screen, or mixing air with powder medicament in a mixing chamber with subsequent introduction of the powder to the patient through the mouthpiece of the device, such as described in Mulhauser, P., et al, U.S. Pat. No. 5,388,572, Sep. 30, 1997, incorporated herein by reference.

An inhaleable PEG-insulin composition may also be delivered using a pressurized, metered dose inhaler (MDI), e.g., the Ventolin.RTM. metered dose inhaler, containing a solution or suspension of drug in a pharmaceutically inert liquid propellant, e.g., a chlorofluorocarbon or fluorocarbon, as described in Laube, et al., U.S. Pat. No. 5,320,094, Jun. 14, 1994, and in Rubsamen, R. M., et al, U.S. Pat. No. 5,672,581 (1994), both incorporated herein by reference.

Alternatively, the PEG-insulins described herein may be dissolved or suspended in a solvent, e.g., water or saline, and administered by nebulization. Nebulizers for delivering an aerosolized solution include the AERx.TM. (Aradigm), the Ultravent.RTM. (Mallinkrodt), the Pari LC Plus.TM. or the Pari LC Star.TM. (Pari GmbH, Germany), the DeVilbiss Pulmo-Aide, and the Acorn II.RTM. (Marquest Medical Products).

As previously described, the PEG-insulin conjugates described herein can also be administered parenterally by intravenous injection, or less preferably by intramuscular or by subcutaneous injection. Precise components of such formulations can be readily determined by one skilled in the art. Suitable formulation types for parenteral administration include ready-for-injection solutions, dry powders for combination with a solvent prior to use, suspensions ready for injection, dry insoluble compositions for combination with a vehicle prior to use, emulsions and liquid concentrates for dilution prior to administration. For instance, an injectable solution of a PEG-insulin composition of the invention may include the composition dissolved in an aqueous vehicle such as aqueous sodium chloride, Ringers solution, a dextrose-injection solution, lactated Ringers solution and the like, and may include one or more pharmaceutically acceptable compatible excipients or additives as described above.

IV. Utility

The compositions of the invention are useful, when administered by any suitable route of administration, and preferably by inhalation or by injection, in a therapeutically effective amount to a mammalian subject, for treating diabetes mellitus, and in particular, type I or type II diabetes.

Claim 1 of 44 Claims

What is claimed is:

1. A dry powder composition for pulmonary administration, said composition comprising a conjugate of insulin covalently coupled to one or more molecules of polyethylene glycol, wherein said powder (i) is characterized by having an emitted dose value of at least about 50% and (ii) when administered to a subject by inhalation, sustains elevated blood levels of insulin in said subject for at least about 6 hours post administration.

____________________________________________
If you want to learn more about this patent, please go directly to the U.S. Patent and Trademark Office Web site to access the full patent.