Particles incorporating surfactants for pulmonary drug delivery

Abstract

Improved aerodynamically light particles for drug delivery to the pulmonary system, and methods for their synthesis and administration are provided. In a preferred embodiment, the aerodynamically light particles are made of a biodegradable material and have a tap density less than 0.4 g/cm³ and a mass mean diameter between 5 μm and 30 μm. The particles may be formed of biodegradable materials such as biodegradable polymers. For example, the particles may be formed of a functionalized polyester graft copolymer consisting of a linear α-hydroxy-acid polyester backbone having at least one amino acid group incorporated therein and at least one poly(amino acid) side chain extending from an amino acid group in the polyester backbone. In one embodiment, aerodynamically light particles having a large mean diameter, for example greater than 5 μm, can be used for enhanced delivery of a therapeutic agent to the alveolar region of the lung. The aerodynamically light particles incorporating a therapeutic agent may be effectively aerosolized for administration to the respiratory tract to permit systemic or local delivery of wide variety of therapeutic agents.

Description

The government has certain rights in this invention by virtue of Grant Number HD29129 awarded to the National Institutes of Health to Robert S. Langer.

08/739,308, filed Oct. 29, 1996, the disclosure of which is incorporated herein.

Aerodynamically Light Particle Size

The mass mean diameter of the particles can be measured using a Coulter Multisizer II (Coulter Electronics, Luton, Beds, England). The aerodynamically light particles in one preferred embodiment are at least about 5 microns in diameter. The diameter of particles in a sample will range depending upon factors such as particle composition and methods of synthesis. The distribution of size of particles in a sample can be selected to permit optimal deposition within targeted sites within the respiratory tract.

The aerodynamically light particles may be fabricated or separated, for example by filtration or centrifugation, to provide a particle sample with a preselected size distribution. For example, greater than 30%, 50%, 70%, or 80% of the particles in a sample can have a diameter within a selected range of at least 5 μm. The selected range within which a certain percentage of the particles must fall may be for example, between about 5 and 30 μm, or optionally between 5 and 15 μm. In one preferred embodiment, at least a portion of the particles have a diameter between about 9 and 11 μm. Optionally, the particle sample also can be fabricated wherein at least 90%, or optionally 95% or 99%, have a diameter within the selected range. The presence of the higher proportion of the aerodynamically light, larger diameter (at least about 5 μm) particles in the particle sample enhances the delivery of therapeutic or diagnostic agents incorporated therein to the deep lung.

In one embodiment, in the particle sample, the interquartile range may be 2 μm, with a mean diameter for example, between about 7.5 and 13.5 μm. Thus, for example, between at least 30% and 40% of the particles may have diameters within the selected range. Preferably, the percentages of particles have diameters within a 1 μm range, for example, 6.0-7.0 μm, 10.0-11.0 μm or 13.0-14.0 μm.

The aerodynamically light particles incorporating a therapeutic drug, and having a tap density less than about 0.4 g/cm³, with mean diameters of at least about 5 μm, are more capable of escaping inertial and gravitational deposition in the oropharyngeal region, and are targeted to the airways or the deep lung. The use of larger particles (mean diameter at least about 5 μm) is advantageous since they are able to aerosolize more efficiently than smaller, non-light aerosol particles such as those currently used for inhalation therapies.

In comparison to smaller non-light particles, the larger (at least 5 μm) aerodynamically light particles also can potentially more successfully avoid phagocytic engulfment by alveolar macrophages and clearance from the lungs, due to size exclusion of the particles from the phagocytes' cytosolic space. Phagocytosis of particles by alveolar macrophages diminishes precipitously as particle diameter increases beyond 3 μm. Kawaguchi, H. et al., Biomatetials 7: 61-66 (1986); Krenis, L. J. and Strauss, B., Proc. Soc. Exp. Med., 107:748-750 (1961); and Rudt, S. and Muller, R. H., J. Contr. Rel., 22: 263-272 (1992). For particles of statistically isotropic shape, such as spheres with rough surfaces, the particle envelope volume is approximately equivalent to the volume of cytosolic space required within a macrophage for complete particle phagocytosis.

Aerodynamically light particles thus are capable of a longer term release of a therapeutic agent in the lungs. Following inhalation, aerodynamically light biodegradable particles can deposit in the lungs (due to their relatively low tap density), and subsequently undergo slow degradation and drug release, without the majority of the particles being phagocytosed by alveolar macrophages. The drug can be delivered relatively slowly into the alveolar fluid, and at a controlled rate into the blood stream, minimizing possible toxic responses of exposed cells to an excessively high concentration of the drug. The aerodynamically light particles thus are highly suitable for inhalation therapies, particularly in controlled release applications.

The preferred mean diameter for aerodynamically light particles for inhalation is at least about 5 μm, for example between about 5 and 30 μm. The particles may be fabricated with the appropriate material, surface roughness, diameter and tap density for localized delivery to selected regions of the respiratory tract such as the deep lung or upper airways. For example, higher density or larger particles may be used for upper airway delivery, or a mixture of different sized particles in a sample, provided with the same or different therapeutic agent may be administered to target different regions of the lung in one administration.

Density and Deposition of Aerodynamically Light Particles

As used herein, the phrase "aerodynamically light particles" refers to particles having a tap density less than about 0.4 g/cm³. The tap density of particles of a dry powder may be obtained using a GeoPyc™ (Micrometrics Instrument Corp., Norcross, Ga. 30093). Tap density is a standard measure of the envelope mass density. The envelope mass density of an isotropic particle is defined as the mass of the particle divided by the minimum sphere envelope volume within which it can be enclosed. Features which can contribute to low tap density include irregular surface texture and porous structure.

Inertial impaction and gravitational settling of aerosols are predominant deposition mechanisms in the airways and acini of the lungs during normal breathing conditions. Edwards, D. A., J. Aerosol Sci., 26: 293-317 (1995). The importance of both deposition mechanisms increases in proportion to the mass of aerosols and not to particle (or envelope) volume. Since the site of aerosol deposition in the lungs is determined by the mass of the aerosol (at least for particles of mean aerodynamic diameter greater than approximately 1 μm), diminishing the tap density by increasing particle surface irregularities and particle porosity permits the delivery of larger particle envelope volumes into the lungs, all other physical parameters being equal.

The low tap density particles have a small aerodynamic diameter in comparison to the actual envelope sphere diameter. The aerodynamic diameter, d_aer, is related to the envelope sphere diameter, d (Gonda, I., "Physico-chemical principles in aerosol delivery," in Topics in Pharmaceutical Sciences 1991 (eds. D. J. A. Crommelin and K. K. Midha), pp. 95-117, Stuttgart: Medpharm Scientific Publishers, 1992)), by the formula: ##EQU1##

where the envelope mass ρ is in units of g/cm³. Maximal deposition of monodisperse aerosol particles in the alveolar region of the human lung (∼60%) occurs for an aerodynamic diameter of approximately d_aer =3 μm. Heyder, J. et al., J. Aerosol Sci., 17: 811-825 (1986). Due to their small envelope mass density, the actual diameter d of aerodynamically light particles comprising a monodisperse inhaled powder that will exhibit maximum deep-lung deposition is: ##EQU2##

where d is always greater than 3 μm. For example, aerodynamically light particles that display an envelope mass density, ρ=0.1 g/cm³, will exhibit a maximum deposition for particles having envelope diameters as large as 9.5 μm. The increased particle size diminishes interparticle adhesion forces. Visser, J., Powder Technology, 58:1-10. Thus, large particle size increases efficiency of aerosolization to the deep lung for particles of low envelope mass density, in addition to contributing to lower phagocytic losses.

Targeting of Particles

Targeting molecules can be attached to the particles via reactive functional groups on the particles. For example, targeting molecules can be attached to the amino acid groups of functionalized polyester graft copolymer particles, such as PLAL-Lys particles. Targeting molecules permit binding interaction of the particle with specific receptor sites, such as those within the lungs. The particles can be targeted by attachment of ligands which specifically or non-specifically bind to particular targets. Exemplary targeting molecules include antibodies and fragments thereof including the variable regions, lectins, and hormones or other organic molecules capable of specific binding, for example, to receptors on the surfaces of the target cells.

Therapeutic Agents

Any of a variety of therapeutic, prophylactic or diagnostic agents can be incorporated within the particles, or used to prepare particles consisting solely of the agent and surfactant. The particles can be used to locally or systemically deliver a variety of therapeutic agents to an animal. Examples include synthetic inorganic and organic compounds, proteins and peptides, polysaccharides and other sugars, lipids, and DNA and RNA nucleic acid sequences having therapeutic, prophylactic or diagnostic activities. Nucleic acid sequences include genes, antisense molecules which bind to complementary DNA to inhibit transcription, and ribozymes. The agents to be incorporated can have a variety of biological activities, such as vasoactive agents, neuroactive agents, hormones, anticoagulants, immunomodulating agents, cytotoxic agents, prophylactic agents, antibiotics, antivirals, antisense, antigens, and antibodies. In some instances, the proteins may be antibodies or antigens which otherwise would have to be administered by injection to elicit an appropriate response. Compounds with a wide range of molecular weight can be encapsulated, for example, between 100 and 500,000 grams or more per mole.

Proteins are defined as consisting of 100 amino acid residues or more; peptides are less than 100 amino acid residues. Unless otherwise stated, the term protein refers to both proteins and peptides. Examples include insulin and other hormones. Polysaccharides, such as heparin, can also be administered.

The polymeric aerosols are useful as carriers for a variety of inhalation therapies. They can be used to encapsulate small and large drugs, release encapsulated drugs over time periods ranging from hours to months, and withstand extreme conditions during aerosolization or following deposition in the lungs that might otherwise harm the encapsulated therapeutic.

The particles may include a therapeutic agent for local delivery within the lung, such as agents for the treatment of asthma, emphysema, or cystic fibrosis, or for systemic treatment. For example, genes for the treatment of diseases such as cystic fibrosis can be administered, as can beta agonists for asthma. Other specific therapeutic agents include, but are not limited to, insulin, calcitonin, leuprolide (or gonadotropin-releasing hormone ("LHRH")), granulocyte colony-stimulating factor ("G-CSF"), parathyroid hormone-related peptide, somatostatin, testosterone, progesterone, estradiol, nicotine, fentanyl, norethisterone, clonidine, scopolomine, salicylate, cromolyn sodium, salmeterol, formeterol, albuterol, and vallium.

Administration

The particles incorporating a surfactant and a therapeutic agent may be administered alone or in any appropriate pharmaceutical carrier, such as a liquid, for example saline, or a powder, for administration to the respiratory system. They can be co-delivered with larger carrier particles, not including a therapeutic agent, the latter possessing mass mean diameters for example in the range 50 μm-100 μm.

Aerosol dosage, formulations and delivery systems may be selected for a particular therapeutic application, as described, for example, in Gonda, I. "Aerosols for delivery of therapeutic and diagnostic agents to the respiratory tract," in Critical Reviews in Therapeutic Drug Carrier Systems, 6:273-313, 1990; and in Moren, "Aerosol dosage forms and formulations," in: Aerosols in Medicine, Principles, Diagnosis and Therapy, Moren, et al., Eds. Esevier, Amsterdam, 1985, the disclosures of which are incorporated herein by reference.

The greater efficiency of aerosolization by particles incorporating a surfactant permits more drug to be delivered. The use of biodegradable polymers permits controlled release in the lungs and long-time local action or systemic bioavailability. Denaturation of macromolecular drugs can be minimized during aerosolization since macromolecules are contained and protected within a polymeric shell. Coencapsulation of peptides with peptidase-inhibitors can minimize peptide enzymatic degradation. Pulmonary delivery advantageously can eliminate the need for injection. For example, the requirement for daily insulin injections can be avoided.

The present invention will be further understood by reference to the following non-limiting examples.

Example 1

PAC Synthesis of Aerodynamically Light Poly[(p-carboxyphenoxy)-hexane anhydride] ("PCPH") Particles

Aerodynamically light poly[(p-carboxyphenoxy)-hexane anhydride] ("PCPH") particles were synthesized as follows. 100 mg PCPH (MW∼25,000) was dissolved in 3.0 mL methylene chloride. To this clear solution was added 5.0 mL 1% w/v aqueous polyvinyl alcohol (PVA, MW ∼25,000, 88 mole % hydrolyzed) saturated with methylene chloride, and the mixture was vortexed (Vortex Genie 2, Fisher Scientific) at maximum speed for one minute. The resulting milky-white emulsion was poured into a beaker containing 95 mL 1% PVA and homogenized (Silverson Homogenizers) at 6000 RPM for one minute using a 0.75 inch tip. After homogenization, the mixture was stirred with a magnetic stirring bar and the methylene chloride quickly extracted from the polymer particles by adding 2 mL isopropyl alcohol. The mixture was continued to stir for 35 minutes to allow complete hardening of the microparticles. The hardened particles were collected by centrifugation and washed several times with double distilled water. The particles were freeze dried to obtain a free-flowing powder void of clumps. Yield, 85-90%.

The mean diameter of a typical batch prepared by this protocol is 6.0 μm, however, particles with mean diameters ranging from a few hundred nanometers to several millimeters may be made with only slight modifications. Scanning electron micrograph photos of a typical batch of PCPH particles showed the particles to be highly porous with irregular surface shape. The particles have a tap density less than 0.4 g/cm³.

A surfactant such as DPPC may be incorporated into the polymer solution prior to particle formation, or optionally the particles can be ionically or covalently coated by surfactant on the particle surface after particle formation, or the surfactant may be absorbed onto the particle surface.

Example 2

PAC Synthesis of Spray-Dried Particles

Aerodynamically Light Particles Containing Polymer and Drug Soluble in Common Solvent

Aerodynamically light 50:50 PLGA particles were prepared by spray drying with testosterone encapsulated within the particles according to the following procedures. 2.0 g poly (D,L-lactic-co-glycolic acid) with a molar ratio of 50:50 (PLGA 50:50, Resomer RG503, B.I. Chemicals, Montvale, N.J.) and 0.50 g testosterone (Sigma Chemical Co., St. Louis, Mo.) are completely dissolved in 100 mL dichloromethane at room temperature. The mixture is subsequently spray-dried through a 0.5 mm nozzle at a flow rate of 5 mL/min using a Buchi laboratory spray-drier (model 190, Buchi, Germany). The flow rate of compressed air is 700 nl. The inlet temperature is set to 30°C and the outlet temperature to 25°C The aspirator is set to achieve a vacuum of -20 to -25 bar. The yield is 51% and the mean particle size is approximately 5 μm. Larger particle size can be achieved by lowering the inlet compressed air flow rate, as well as by changing other variables. The particles are aerodynamically light, as determined by a tap density less than or equal to 0.4 g/cm³. Porosity and surface roughness can be increased by varying the inlet and outlet temperatures, among other factors.

Aerodynamically Light Particles Containing Polymer and Drug in Different Solvents

Aerodynamically light PLA particles with a model hydrophilic drug (dextran) were prepared by spray drying using the following procedure. 2.0 mL of an aqueous 10% w/v FITC-dextran (MW 70,000, Sigma Chemical Co.) solution was emulsified into 100 mL of a 2% w/v solution of poly (D,L-lactic acid) (PLA, Resomer R206, B.I. Chemicals) in dichloromethane by probe sonication (Sonics & Materials, Model VC-250 sonicator, Danbury, Conn.). The emulsion is subsequently spray-dried at a flow rate of 5 mL/min with an air flow rate of 700 nl/h (inlet temperature=30°C, outlet temperature=21°C, -20 mbar vacuum). The yield is 56%. The particles are aerodynamically light, as determined by a tap density less 0.4 g/cm³.

Aerodynamically Light Protein Particles

Aerodynamically light lysozyme particles were prepared by spray drying using the following procedure. 4.75 g lysozyme (Sigma) was dissolved in 95 mL double distilled water (5% w/v solution) and spray-dried using a 0.5 mm nozzle and a Buchi laboratory spray-drier. The flow rate of compressed air was 725 nl/h. The flow rate of the lysozyme solution was set such that, at a set inlet temperature of 97°-100°C, the outlet temperature is 55°-57°C The aspirator was set to achieve a vacuum of -30 mbar. The enzymatic activity of lysozyme was found to be unaffected by this process and the yield of the aerodynamically light particles (tap density less than 0.4 g/cm³) was 66%.

Aerodynamically Light High-Molecular Weight Water-Soluble Particles

Aerodynamically light dextran particles were prepared by spray drying using the following procedure. 6.04 g DEAE dextran (Sigma) was dissolved in 242 mL double distilled water (2.5% w/v solution) and spray-dried using a 0.5 mm nozzle and a Buchi laboratory spray-drier. The flow rate of compressed air was 750 nl/h. The flow rate of the DEAE-dextran solution was set such that, at a set inlet temperature of 155°C, the outlet temperature was 80°C The aspirator was set to achieve a vacuum of -20 mbar. The yield of the aerodynamically light particles (tap density less than 0.4 g/cm³) was 66%.

Aerodynamically Light Low-Molecular Weight Water-Soluble Particles

Aerodynamically light trehalose particles were prepared by spray drying using the following procedure. 4.9 g trehalose (Sigma) was dissolved in 192 mL double distilled water (2.5% w/v solution) and spray-dried using a 0.5 mm nozzle and a Buchi laboratory spray-drier. The flow rate of compressed air 650 nl/h. The flow rate of the trehalose solution was set such that, at a set inlet temperature of 100°C, the outlet temperature was 60°C The aspirator was set to achieve a vacuum of -30 mbar. The yield of the aerodynamically light particles (tap density less than 0.4 g/cm³) was 36%.

Aerodynamically Light Low-Molecular Weight Water-Soluble Particles

Polyethylene glycol (PEG) is a water-soluble macromolecule, however, it cannot be spray dried from an aqueous solution since it melts at room temperatures below that needed to evaporate water. As a result, PEG was spray-dried at low temperatures from a solution in dichloromethane, a low-boiling organic solvent. Aerodynamically light PEG particles were prepared spray drying using the following procedure. 5.0 g PEG (MW 15,000-20,000, Sigma) was dissolved in 100 mL double distilled water (5.0% w/v solution) and spray-dried using a 0.5 mm nozzle and a Buchi laboratory spray-drier. The flow rate of compressed air 750 nl/h. The flow rate of the PEG solution was set such that, at a set inlet temperature of 45°C, the outlet temperature was 34°-35°C The aspirator was set to achieve a vacuum of -22 mbar. The yield of the aerodynamically light particles (tap density less than 0.4 g/cm³) was 67%.

A surfactant such as DPPC may be incorporated into the polymer solution prior to particle formation, or optionally the particles can be ionically or covalently coated by surfactant on the particle surface after particle formation, or the surfactant may be absorbed onto the particle surface.

Materials and Methods

In Examples 3 and 4 below, the following materials and methods were used.

Materials

The polymers: poly(D,L-lactic-co-glycolic acid) with a molar ratio of 50:50 and reported molecular weights of 100,000 Daltons (PLGA RG506) and 34,000 Daltons (PLGA RG503), and poly(D,L-lactic acid) with a reported molecular weight of 100,000 Daltons (PLA R206) were obtained from Boehringer Ingelheim (distributed by B.I. Chemicals, Montvale, N.J.). Fluorescently labelled FITC-Dextran with an average molecular weight of 19,000, and L,α-phosphatidylcholine dipalmitoyl (DPPC) were purchased from Sigma Chemical Company, St. Louis, Mo.

Microsphere Preparation: Double Emulsion

A double-emulsion solvent-evaporation procedure (Cohen, S., et al., Pharm. Res., 8(6): 713-720 (1991); and Tabata, Y., et al., Pharm. Res., 10(4): 487-496 (1993)), was modified to prepare microspheres for aerosolization. Briefly, 300 μl of an aqueous FITC-Dextran solution (50 mg/ml) was emulsified on ice into a 4.0 ml polymer solution is methylene chloride (200 mg polymer) by sonication at output 3 (Model VC-250, Sonics & Materials Inc., Danbury, Conn.) using a microtip for 5-10 s to form the inner-emulsion. The first emulsion was poured into 100 ml 1.0% aqueous PVA solution and homogenized (Model LD4 Homogenizer, Silverson Machines Ltd, England) at 6000 RPM using a 5/8" tip for 1 min to form the double emulsion. The microspheres were continuously stirred for 3 hours to allow hardening, collected by centrifugation, washed several times with double-distilled water, and freeze-dried into a freely flowing powder. Microspheres containing DPPC were prepared by dissolving DPPC in the polymer solution at 3 mg/ml prior to the initial emulsification.

Microsphere Preparation: Spray Drying

The model hydrophilic drug, dextran labeled with fluorescein isothiocynate (FITC-Dextran), was encapsulated into PLA or PLGA by a novel emulsion/spray method. For example, 2.0 ml of an aqueous 10% w/v FITC-Dextran (MW=70,000, Sigma Chemical Co.) solution was emulsified into 100 ml of a 2% w/v solution of PLA in dichloromethane by probe sonication. The emulsion was subsequently spray-dried using a Buchi Mini Spray Drier (Model 190, Buchi Instruments, Germany) at a flow rate of 5 ml/min with an inlet air flow rate of 700 nl/h, inlet temperature of 30°C, outlet temperature of 21°C, and vacuum of -20 mbar. When DPPC was incorporated it was dissolved in the polymer solution at a concentration of 2 mg/ml prior to emulsification and spray drying.

Microsphere Size Distribution Analysis

Microsphere size distributions were determined using a Coulter Multisizer II (Coulter Electronics Limited, Luton, Beds, England). Approximately 10 drops Coulter type IA non-ionic dispersant were added, followed by 2 mL isoton II solution (Coulter), to 5-10 mg microspheres, and the spheres were dispersed by brief vortex mixing. This suspension was added to 50 mL isoton 11 solution until the coincidence of particles was between 5 and 8%. Greater than 500,000 particles were counted for each batch of spheres.

Drug Distribution by Confocal Microscopy

For confocal microscopy, a few milligrams of microspheres containing FITC-Dextran as the drug were suspended in glycerin by brief probe sonication (Vibra-cell Model VC-250 Sonicator, 1/8" microtip probe, Sonics & Materials Inc., Danbury, Conn.) at output 4 (50 W). A drop of the suspension was placed onto a glass slide and a glass cover slip was applied and held in place with finger nail polish. The suspension was allowed to settle for one hour before being viewed by confocal microscopy (Bio-Rad MRC-600 Confocal, Axioplan microscope).

Microsphere Morphology by Scanning Electron Microscopy (SEM)

Microsphere morphology was observed by scanning electron microscopy (SEM) using a Stereoscan 250 MK3 microscope from Cambridge Instruments (Cambridge, Mass.) at 15 kV. Microspheres were freeze-dried, mounted on metal stubs with double-sided tape, and coated with gold prior to observation.

Microsphere Density Analysis

Microsphere bulk density was estimated by tap density measurements and confirmed by mercury intrusion analysis at Porous Materials, Inc. (Ithaca, N.Y.).

Determination of Amount FITC-Dextran and DPPC Encapsulated

The amount of model drug, FITC-Dextran, encapsulated into microspheres was determined by dissolving 10.0 mg microspheres in 3.0 ml 0.8N NaOH overnight at 37°C, filtering with a 0.45 μm filter (Millipore), and measuring the fluorescence relative to a standard curve (494 nm excitation and 525 nm emission) using a fluorimeter. The drug loading was determined by dividing the amount of FITC-Dextran encapsulated by the theoretical amount if it all were encapsulated. The amount of lung surfactant, DPPC, encapsulated into microspheres was determined by dissolving 10.0 mg of microspheres in chloroform and using the Stewart Assay. New, R. R. C., "Characterization of Liposomes," in Liposomes: A Practical Approach, R. New, Editor, IRL Press, New York, 105-161 (1990).

In Vitro Aerosolization and Inertial Deposition Behavior

The in vitro microparticle aerodynamic characteristics were studied using an Andersen Mark I Cascade Impactor (Andersen Samplers, Atlanta, Ga.) at an air flow rate of 28.3 l/min. The metal impaction plates were coated with a thin film of Tween 80 minimize particle bouncing Turner, J. and S. Hering, J. Aerosol Sci., 18: 215-224 (1987). Gelatin capsules (Eli Lilly) were charged with 20 mg of microparticles and loaded into a Spinhaler® inhalation device (Fisons, Bedford, Mass.). The aerosolization experiments were done in triplicate. In each experiment, 10 inhalers were discharged for 30 seconds into the impactor. A 60-second interval was observed between every two consecutive aerosolizations. Fractions of microspheres deposited on each of nine stages, corresponding to stages 0-7, and the filter (F) of the impactor, were collected in volumetric flasks by carefully washing the plates with NaOH solution (0.8N) in order to provide degradation of the polymer and complete dissolution of the fluorescent material. After 12 hours of incubation at 37°C, the solutions were filtered with a 0.45 μm filter and the amount of fluorescent material in each stage was measured at 494 nm (excitation) and 525 nm (emission) using a fluorimeter. Respirable fraction of the delivered dose was calculated according to the fluorescence measurements as percentages of the total fluorescence (i.e., that amount collected in stages 0 - Filter) compared with that collected in stages 2 - Filter of the Impactor.

In Vivo Particle Distribution Following Aerosolization in Rats

Male Sprague Dawley rats (150-200 g) were anesthetized using a mixture of ketamine (90 mg/kg) and xylazine (10 mg/kg). The anesthetized rat was placed ventral side up on a surgical table provided with a temperature controlled pad to maintain physiological temperature. The animal was cannulated above the carina with an endotracheal tube connected to a Harvard ventilator (Rodent Ventilator Model 685, South Natick, Mass.). The animal was force ventilated for 20 minutes at 300 ml/min. 50 mg of microspheres made with or without DPPC were introduced into the endotracheal tube. Following the period of forced ventilation, the animal was sacrificed and the lungs and trachea were separately washed using broncholalveolar lavage as follows: a tracheal cannula was inserted, tied into placed, and the airways were washed with 10 ml aliquots of phenol red-free Hanks balanced salt solution (Gibco, Grand Island, N.Y.) without Ca²+ and Mg²+ (HBSS). The lavage procedure was repeated until a total volume of 30 ml was collected. The lavage fluid was centrifuged (400 g) and the pellets collected and resuspended in 2 ml HBSS. 100 μl was removed for particle counting using a hemacytometer. The remaining solution was mixed with 10 ml of 0.4N NaOH. After incubation at 37° C. for 12 hours, the fluorescence of each solution was measured (wavelengths of 494 nm excitation, 525 nm emission) using a fluorimeter.

Example 3

Fabrication of PLGA microspheres by a Double Emulsion Process Which Encapsulate a Model High-Molecular-Weight Drug, FITC-Dextran.

Scanning electron microscopy "SEM" photographs showing surface morphology of microspheres (MS) made by the double emulsion process with and without the lung surfactant, DPPC were obtained. By SEM, the microspheres made with and without DPPC by the double emulsion process had very similar surface characteristics and size distribution, as confirmed by size distribution measurements, shown below in Table 1.

The efficient entrapment of DPPC within microspheres (83% of theoretical ±11% standard deviation, n=6) was confirmed by dissolving an aliquot of MS in chloroform and detecting the DPPC concentration in solution by the Stewart Assay, as shown in Table 1. Particles made by double emulsion with DPPC are easily resuspended in aqueous solution after lyophilization and are lump-free when dry as determined by light microscopy. Particles made by the double emulsion process without DPPC resuspend easily, however, they appear somewhat agglomerated when dry by light microscopy.

TABLE 1
Characteristics of Microparticles used for In Vitro and In
Vivo Aerosolizationa
Mass-Mean                             FITC-Dextran
(True)      DPPC Load                 (Model Drug)
Diameter,   (μg/mg  DPPC Loading  Loading
Sample    (μm)    spheres)    Efficiency, (%) Efficiency, (%)
MS        8.5 ± 0.76 0           N/A           95.8
without
DPPC
MS with   8.2 ± 0.18 45 ± 6 83 ± 11 82.4
DPPC
1 Values are given ± standard deviation.

Confocal microscopy was used to evaluate the distribution of the model drug, FITC-Dextran (M_w 19,000), throughout microspheres made without DPPC and with DPPC. In each case, the drug is evenly dispersed throughout the polymer matrix, which can lead to prolonged delivery of macromolecules after placement in an aqueous environment.

The density of the microspheres as determined by mercury intrusion analysis is shown in Table 2 (and confirmed by tap density measurements).

TABLE 2
Comparison of Porous Microparticles with Bulk
(PLGA 50:50) Polymer
Density, ρMS  Respirable Size
Sample               (g/cc)      Range, dresp (μm)
Bulk PLGA            1.35        0.69-4.05
MS without DPPC      0.37 ± 0.03 1.3-7.7
MS with DPPC         0.30 ± 0.06 1.46-8.58

Using the concept of aerodynamic diameter (Gonda, I., in Topics in Pharmaceutical Sciences 1991, D. Crommelin and K. Midha, Editors, Stuttgart: Medpharm Scientific Publishers, pp. 95-117 (1992)), it is possible to determine the size range of the microspheres which are theoretically respirable given their mass density, ρ_MS. Specifically, it can be shown below in Equation 2 that: ##EQU3##

where d_resp corresponds to the diameter of particles (in μm) theoretically able to enter and remain in the airways without inertial or gravitational deposition (particles smaller than this range are exhaled), and where ρ_MS is in units of g/cc. The theoretical respirable size range of the microspheres also is shown in Table 2. The optimal size range (i.e., d_resp) for a non-porous PLGA 50:50 microsphere is 0.69-4.05 μm (Table 2). The optimal respirable size range for microspheres without DPPC is 1.3-7.7 μm and, for microspheres with DPPC, 1.46-8.58 μm (Table 2). The upper limit on size of respirable particles is increased from 4.05 to greater than 8.5 μm when DPPC is used in the PLGA microsphere preparation. Therefore, the use of low density DPPC microspheres allows the use of larger particles for aerosolization, which may have advantages for drug delivery, such as less particle-particle interaction due to decreased surface area to volume ratio, and lower susceptibility to phagocytosis by alveolar macrophages. In addition, a primary effect of DPPC is to render the particles less adhesive and therefore allow improved aerosolization, as demonstrated below.

FIGS. 1 and 2 show the results of an in vitro aerosolization of the PLGA microspheres made by a double emulsion process with and without DPPC. The microspheres were aerosolized as a dry powder released from a Spinhaler® dry powder inhaler (DPI). FIG. 1 illustrates the mass-fraction of the initial dose that is released from the dry powder inhaler device (DPI Efficiency) using an Andersen Mark I Cascade Impactor. DPI efficiencies approaching 80% were obtained with microspheres made with and without DPPC. Although the DPI efficiencies for the two batches were nearly the same, a great difference can be seen between microspheres made with and without DPPC when their deposition within the cascade impactor is observed (FIG. 2).

FIG. 2 shows the mass fraction of aerosolized particles that is deposited in stages 2 through Filter (2-Filter) of the Andersen cascade impactor, considered the stages corresponding to the respirable fraction of the microspheres. Stages 0 and 1 correspond roughly to the mouth and throat, and to the upper airways of the lung, respectively. Stages 2-F correspond to successively deeper fractions of the lung. It can be seen that a much greater percentage of microspheres make it to the latter stages of the impactor (considered deeper portions of the lungs) when DPPC is used in their preparation. Overall, greater than 35% (37.0±2.1) of aerosolized particles made with DPPC are considered respirable compared with 13.2±2.9% without DPPC, as shown in Table 3. The large difference in respirable fraction between the DPPC and non-DPPC particles is at least in part attributed to reduced particle-particle interaction due to the use of DPPC.

In order to estimate the theoretical respirable fraction (RF) of the microspheres, and compare it with experimentally measured in vitro and in vivo RF's, size distribution measurements were analyzed to determine the percentage of particles (by mass) of each type (DPPC and non-DPPC) that were within the theoretical respirable size range (i.e., d_resp Table 2). As shown in Table 3, a higher percentage of particles made with DPPC are expected to be respirable compared with non-DPPC particles (63 to 51%, respectively). This theoretical respirable fraction is based on the mass fraction of microspheres with diameters in the respirable size range, d_resp as defined by Eq. (2), and therefore takes into account the different sizes and densities of the two batches of microspheres.

TABLE 3
Comparison of Microparticle Aerosolization
Properties In Vitro
Theoretical Respirable
Fraction
(i.e., Mass % of
microspheres in      Measured Respirable
Sample           Respirable Size Range)a  Fraction (%, In Vitrob)
microspheres     51 ± 6        13.2 ± 2.9
without DPPC
microspheres with 63 ± 2        37.0 ± 2.1
DPPC
a Based on theoretical respirable size range (dresp Table 2) and
size distribution analyses.
b Measured using an Andersen Mark I Cascade Impactor.

To determine whether agglomeration forces during particle aerosolization from the Spinhaler device might be playing a role even after the particles enter the impactor system (i.e., primarily non-DPC particles remain agglomerated in the inspired stream, resulting in deposition in the first two impactor stages: stages 0 and 1), in vivo aerosolization experiments were performed in which particles were permitted to fall by gravity into the inspiration stream of a Harvard ventilator system joined with the trachea of an anesthetized rate. In this model, approximately 63% of the inhaled DPPC-PLGA particles deposit in the airways and distal lung regions, whereas 57% of the non-DPPC particles are able to penetrate beyond the trachea in the lungs. These respirable fractions are much nearer to the predicted respirable fractions based upon particle diameter and mass density (Table 3).

Particle aggregation thus is less with DPPC-containing PLGA particles than without DPPC, even though the particles are of similar size and surface morphological features. The use of DPPC thus appears to reduce interparticle attractions, such as van der Waals and electrostatic attractions. It is also possible that the presence of DPPC reduces moisture absorption which may cause particle-particle interaction by capillary forces.

In addition to the biocompatibility features of DPPC and improvement of surface properties of microspheres for aerosolization, it is possible that the release of DPPC from the slow-eroding PLGA microspheres in the alveolar region of the lungs can more effectively insure the maintenance of normal surfactant fluid composition thereby minimizing the possibility of local toxic side effects. The alveolar surfactant fluid layer is, on average, 10 nm thick (Weibel, E. R., Morphometry of the Human Lung, New York: Academic Press (1963).

Example 4

Fabrication of PLGA Microspheres by Spray Drying which Encapsulate a Model High Molecular Weight Drug, FITC-Dextran.

Microspheres were made by spray drying using a variety of polymeric carriers with and without the incorporation of DPPC. The results are summarized in Table 4.

TABLE 4
Characterization of Spray Dried Microparticulates
Mass-     DPPC Load   FITC-       % of
Mean      (μg/mg  Dextran     Surface
(True)    spheres) and Loading     Coated
Diameter, Efficiency  Efficiency, with DPPC
Sample         (μm)  (%)         (%)         by ESCA
R206 + DPPC 5.4       a           54.9        a
R206 - DPPC 4.4       --     64.8        --
RG503 + DPPC 2.0       62.8        65.2        46.5%
RG503 - DPPC 3.0       --     78.2        --
RG506 + DPPC 4.3       89.1        62.7        42-62%
RG506 - DPPC b         --     100         --
a Not Determined
b No reliable determination because the powder was highly aggregated.

Aerosolization properties of the microspheres also were examined, as shown in Table 5. Microspheres made by spray drying with and without DPPC have similar size distributions (Table 5) and mass densities (0.49±0.04 g/cc). However, the aerosolization performance of spray-dried aerosols made with and without DPPC is markedly different. FIG. 3 shows that the fraction of low-molecular-weight PLGA RG503 microparticles that are aerosolized from the dry powder inhaler (i.e., the % of particles that leave the DPI upon simulated inhalation, defined as the DPI Efficiency) is 70.4% when the particles are made with DPPC compared with only 46.8% for particles made without DPPC. Furthermore, the deposition of all types of polymer microparticles following aerosolization into an Andersen impactor is greatly improved using DPPC-coated particles (Table 5). Without the use of DPPC, ≦2% of the particles aerosolized reach the latter stages of the impactor (those corresponding to the respirable fraction, stages 2-Filter). On the other hand, a maximum of 25.6% of DPPC-coated microspheres reach stages 2-Filter, as shown in FIG. 4. Higher respirable fractions may be obtained with particles that contain low molecular weight drugs that are soluble in methylene chloride and therefore do not require the use of water during their preparation.

TABLE 5
Summary of Aerosolization Data of microspheres Prepared
by Spray Drying with or without DPPC
%           %           %
Aerosolized Aerosolized Aerosolized
Particles that Particles that Particles that
reach stages reach stages reach stages DPI
Sample         1 - Filter  2 - Filter  3 - Filter  Efficiency
R206 + DPPC 40.4 ± 8.4 25.6 ± 2.3  18.0 ± 2.7  38.6 ± 3.7
R206 - DPPC   7.4 ± 2.1 1.8 ± 0.5 1.1 ± 0.3 41.0 ± 4.8
RG503 + DPPC 36.0 ± 9.2 14.7 ± 1.53 10.4 ± 0.46 70.4 ± 2.4
RG503 - DPPC   3.3 ± 0.6 2.1 ± 0.3 2.0 ± 0.3 46.8 ± 8.0
RG506 + DPPC 13.7 ± 9.1 7.1 ± 4.1 4.1 ± 2.5 76.6 ± 8.4
RG506  - DPPC   1.8 ± 0.6 1.6 ± 0.6 1.4 ± 0.7 74.0 ± 7.2
R206 = PLA, molecular weight approximately 100,000.
RG503 = PLGA 50:50, molecular weight approximately 34,000.
RG506 = PLGA, molecular weight approximately 100,000.

Modifications and variations of the present invention will be obvious to those skilled in the art from the foregoing detailed description. Such modifications and variations are intended to come within the scope of the following claims.

Claims

1. A particulate composition for drug delivery to the pulmonary system comprising: #5# biodegradable particles incorporating a therapeutic, prophylactic or diagnostic agent and a surfactant, wherein the particles have a tap density less than 0.4 g/cm³ and a mean diameter between 5 μm and 30 μm effective to yield an aerodynamic diameter of the particles of between approximately one and three microns.