PROCESSES FOR MAKING LACTOSE UTILIZING PRE-CLASSIFICATION TECHNIQUES AND PHARMACEUTICAL FORMULATIONS FORMED

THEREFROM

Field of the Invention

The invention generally relates to processes for making lactose and pharmaceutical formulations formed therefrom.

Background of the Invention

In the field of inhalation therapy, it is generally desirable to employ therapeutic molecules having a particle size (i.e., diameter) in the range of 1 to 10μm. Carrier molecules or excipients, such as lactose, for inhaled therapeutic preparations also include significantly larger diameter particles (e.g., 100 to 150μm) that typically do not penetrate into the upper respiratory tract to the same degree as the active ingredient. In general, it is preferable to use a smaller particle size for the lactose or a lactose blend having a defined ratio of coarse and fine lactose.

The lactose particle size and distribution will also, in many instances, significantly influence pharmaceutical and biological properties, such as, for example, bioavailablity. For example, it is well known that coarse lactose in crystalline form has a fair flow rate and good physical stability whereas fine lactose powder, such as that produced by conventional fine grinding or milling, generally lacks good flow properties. Lactose prepared by conventional spray drying either lacks desired flow properties or contains too many large sized lactose crystals.

It is well known that one particular drawback associated with conventional means of producing pharmaceutical grade lactose relates to undesirable variations in particle size, morphology and distribution. Such production methods are particularly problematic in that they often lead to excessive and undesirable variations in the fine particle mass ("FPMass") of pharmaceutical formulations employing such lactose. FPMass is the weight of medicament within a given dose that reaches the desired size airways to be

effective. For example, a desired size may be defined as approximately 1 micron to 10 microns as measured by laser scattering techniques.

Lactose morphology is believed to be another important parameter to control, and it is believed that the degree of surface roughness can influence the interaction between the lactose particle and excipient and as such is now often measured as part of the lactose selection criteria. See e.g., Pharmaceutical Technology Europe April 2004, page 23.

It is possible that two lactose particles may be measured as having the same particle size, but if one is smooth, eg un-milled crystalline lactose, and the other is a rougher-surfaced milled crystal, that these could associate to a different extent with the active and thus impact upon either the initial FPMass performance or the through life stability performance of the product.

Summary of the Invention

The present invention attempts to address the above problems associated with conventional processes of producing lactose, and provides a process of producing lactose possessing reduced levels of variation for both particle size distribution and particle morphology.

In one aspect, the invention provides a process for forming lactose suitable for use in a pharmaceutical formulation and having a predetermined particle size distribution. The process comprises providing a plurality of lactose particles containing no more than 10% w/w of lactose particles having a volume average particle size of about 70 microns or less; milling the plurality of lactose particles to yield a plurality of milled lactose particles with an average particle size, (D50), ranging from about 50 microns to about 100 microns; and then classifying the plurality of milled lactose particles into at least two fractions comprising a fine fraction and a coarse fraction wherein the fine fraction has an average particle size, (D50), ranging from about 3 microns to about 50 microns, and the coarse fraction has an average particle size, (D50), ranging from about 40 microns to about 250 microns. In one embodiment, appropriate amounts of coarse and fine fractions may be combined with at least one medicament to form a pharmaceutical formulation. These and other aspects are encompassed by the present invention.

Brief Description of the Drawings

FIG. 1 illustrates a particle size profile for lactose formed in accordance with Example 1.

FIG. 2 illustrates a particle size profile for lactose formed in accordance with Example 2.

FIG. 3 illustrates a particle size profile for lactose formed in accordance with Example 3.

FIG. 4 illustrates a comparison of a particle size profile for lactose formed in accordance with Example 2 with two conventional production batches.

FIG. 5 illustrates the particles size distribution of the initial crystalline A feed lactose used in Examples 1 , 2 and 3. This shows the varying amounts of lactose in the less than 60 micron region.

FIG. 6 illustrates the particles size distribution of the crystalline A feed lactose after it has been classified to remove the majority of the smaller crystalline lactose particles.

FIG. 7 illustrates the correlation between D10 and the Span of Feed Lactose Batches.

FIG. 8 illustrates the particle size distribution for pre-classified lactose, coarse fraction.

Detailed Description of the Invention

The invention will now be described with respect to the embodiments set forth herein. It should be appreciated that these embodiments are set forth to illustrate the invention, and that the invention is not limited to these embodiments.

All publications, patents, and patent applications cited herein, whether supra or infra, are hereby incorporated herein by reference in their entirety to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference.

It must be noted that, as used in the specification and appended claims, the singular forms "a", "an", "one" and "the" may include plural references unless the content clearly dictates otherwise.

As used herein, the term "D50" is defined as the size in microns above or below which 50 percent of the particles reside on a volume basis.

The process for forming lactose in accordance with the present invention may encompass various embodiments. For example, in one embodiment, the step of milling the plurality of lactose particles to yield a plurality of lactose particles may encompass obtaining those particles with an average particle size (D50) ranging from about 50, 55, 60, 65, 70 or 75 microns to about 70, 75, 80, 85, 90, 95 or 100 microns. In one embodiment, for example, the step of classifying the plurality of milled lactose particles into at least two fractions may result in a fine fraction having an average particle size (D50) ranging from about 3, 10, 15, 20, 25, 30 or 35 microns to about 30, 35, 40, 45 or 50 microns as well as a coarse fraction having an average particle size (D50) ranging from about 40, 75, 100, 125 or 150 microns to about 100, 125, 150, 175, 200, 225 or 250 microns, as measured by Malvern particle sizing.

In accordance with the present invention, the term "lactose" as used herein is to be broadly construed. As an example, lactose is intended to encompass physical, crystalline, amorphous and polymorphic forms of lactose, including, but not limited to, the stereoisomers α-lactose monohydrate and β-anhydrous lactose, as well as α-anhydrous lactose. Combinations of the above may be used.

Lactose (i.e., milk sugar) is preferably obtained from cheese whey, which can be manufactured in different forms depending on the process employed. As used herein, the term "particle" is to be broadly interpreted to encompass those of various shapes, sizes, and/or textures which can include those that may have varying degrees of irregularities, disuniformities, etc. or which may possess regular and/or uniform properties.

The plurality of lactose particles used in the process of the present invention may be acquired in various manners. In one embodiment, the plurality of lactose particles is present in the form of crystalline or unmilled lactose. In one embodiment, the initial step of providing lactose particles containing no more than 10% w/w of lactose particles having a volume average particle size of about 70 microns or less, comprises obtaining the plurality of lactose particles from a crystallization process. One example of a

suitable crystallization process that may be employed is set forth in copending U.S. Application entitled "Process for Crystallizing Lactose Particles for Use in Pharmaceutical Formulations", Serial No.60/651 ,754 filed concurrently herewith. It should be understood that other processes may also be employed.

In another embodiment, the initial step of providing a plurality of lactose particles containing no more than 10% w/w of lactose particles having a volume average particle size of about 70 microns or less, comprises obtaining the plurality of lactose particles by sieving a source of lactose to produce the plurality of lactose particles. Typical examples of commercially available sieving apparatus are made available by Russell Finex of Charlotte, North Carolina and Alpine Sieves of Augsburg, Germany.

In another embodiment, the plurality of lactose particles containing no more than 10% w/w of lactose particles having a volume average particle size of about 70 microns or less may be obtained by classifying a source of lactose into two fractions comprising the plurality of lactose particles containing no more than 10% w/w of lactose particles having a volume average particle size of about 70 microns or less and a remaining fine fraction. One example of a commercially preferred classifier made commercially available by Hosakawa of Cheshire, United Kingdom. The remaining fine fraction of lactose is lactose that is unmilled or crystalline. Optionally, prior to classifying, the particles may be subjected to sieving.

FIG. 5 illustrates the full particle size profile of the initial feed batches before the smaller crystals have been removed by classification. FIG. 6 illustrates the full particle size profile after the smaller crystals have been removed via a classification process. This results in a much more uniform feed material being presented to the mill.

The Span is often a parameter used to describe the narrowness / broadness of a distribution of particles. The Span is given by the following formulation:

(D90 - D10)/D50 = Span

The term "D90" is defined as the size in microns below which 90 percent of the particles reside on a volume basis.

The term "D10" is defined as the size in microns below which 10 percent of the particles reside on a volume basis.

Although not intending to be bound by theory, for particles with similar D50's the smaller the Span, the tighter are the distribution of particles. In general, feed lactose batches with small Span values would be better input materials to the mill as this would indicate a narrower particle size distribution. It can be seen from FIG. 7 that the higher the D10 values are for typical crystalline A Feed lactose batches, the smaller the Span values. The following Tables 1-2 illustrate that tighter particle size distributions are obtained after the classification of the crystalline A Feed lactose, and hence produce a preferred input into the mill

Table 1

The step of milling the plurality of lactose particles to yield a plurality of milled lactose particles may be carried out using known techniques. As an example, in one embodiment, milling may be performed using an impact mill (e.g., air classifier mill, (ACM)) whereby size reduction of pharmaceutical grade lactose is a combination of milling and classification. In one

embodiment, milling is conducted using a standard air classifier mill fitted with mill rotor and grinding pins, classifier wheel and grinding track. One example of a commercially preferred mill is the MikroPul ACM made commercially available by Hosakawa.

The process described herein is suitable for providing lactose for use in a wide range of inhalation applications, encompassing those that require very fine lactose to those requiring much coarser lactose. Such that in one embodiment the coarse fraction alone may be suitable for pharmaceutical applications and at the other extreme the fine fraction alone may be suitable for pharmaceutical applications. Required FPMass performance from inhalation devices, including, in one embodiment and without limitation Advair, can be targeted by the combination of appropriate amounts of the coarse and fine fractions. This can cover the complete range from 100 percent coarse fraction + 0 percent fine fraction to 100 percent fine fraction + 0 percent coarse fraction. The exact proportions that are required will depend on the particle size profile of each of the coarse and fine fractions themselves.

As an example, the above combining step may be achieved by blending, although other procedures may be employed. A typical blender used would be of the orbital screw type such as the Vrieco-Nauta Conical Blender made commercially available by Hosakawa.

In another aspect, the invention may encompass pharmaceutical formulations formed according to various processes set forth herein.

Medicaments, for the purposes of the invention, include a variety of pharmaceutically active ingredients, such as, for example, those which are useful in inhalation therapy. In general, the term "medicament" is to be broadly construed and include, without limitation, actives, drugs and bioactive agents, as well as biopharmaceuticals. Various embodiments may include medicament present in micronized form. Appropriate medicaments may thus be selected from, for example, analgesics, (e.g., codeine, dihydromorphine, ergotamine, fentanyl or morphine); anginal preparations, (e.g., diltiazem); antiallergics, e.g., cromoglicate, ketotifen or nedocromil); antiinfectives (e.g., cephalosporins, penicillins, streptomycin, sulphonamides, tetracyclines and pentamidine); antihistamines, (e.g., methapyrilene); antiinflammatories, (e.g., beclometasone dipropionate, fluticasone propionate, flunisolide, budesonide,

rofleponide, mometasone furoate, ciclesonide, triamcinolone acetonide, 6α, 9α-difluoro-11 β-hydroxy-16α-methyl-3-oxo-17α-propionyloxy-androsta-1 ,4- diene-17β-carbothioic acid S-(2-oxo-tetrahydro-furan-3-yl) ester), (6α, 11 β, 16α, 17β)-6,9-difluoro-17-{[(fluoromethyl)thio]carbonyl}-11 -hydroxy-16- methyl-3-oxoandrosta-1 ,4-dien-17-yl 2-furoate, and (6α,11β,16α,17α)~6,9- difluoro-17-{[(fluoromethyl)thio]carbonyl}-11 -hydroxy-16-methyl-3- oxoandrosta-1 ,4-dien-17-yl 4-methyl-1 ,3-thiazole-5-carboxylate); antitussives, (e.g., noscapine); bronchodilators, e.g., albuterol (e.g. as sulphate), salmeterol (e.g. as xinafoate), ephedrine, adrenaline, fenoterol (e.g as hydrobromide), formoterol (e.g., as fumarate), isoprenaline, metaproterenol, phenylephrine, phenylpropanolamine, pirbuterol (e.g., as acetate), reproterol (e.g., as hydrochloride), rimiterol, terbutaline (e.g., as sulphate), isoetharine, tulobuterol,4-hydroxy-7-[2-[[2-[[3-(2-(henylethoxy)propyl]su lfonyl]ethyl]- amino]ethyl-2(3H)-benzothiazolone), 3-(4-{[6-({(2R)-2-hydroxy-2-[4-hydroxy-3- (hydroxymethyl) phenyl]ethyl}amino)hexyl]oxy}butyl) benzenesulfonamide, 3- (3-{[7-({(2R)-2-hydroxy-2-[4-hydroxy-3-

(hydroxymethyl)phenyl]ethyl}amino)heptyl] oxyjpropyl) benzenesulfonamide, 4-{(1 /?)-2-[(6-{2-[(2,6-dichlorobenzyl)oxy] ethoxy}hexyl)amino]-1 - hydroxyethyl}-2-(hydroxymethyl)phenol, 2-hydroxy-5-((1 R)- 1 -hydroxy-2-{[2-(4- {[(2R)-2-hydroxy-2-ρhenylethyl]amino} phenyl)ethyl]amino}ethyl)phenylformamide, 8-hydroxy-5-{(1 /?)-1 -hydroxy-2- [(2-{4-[(6-methoxy-1 ,1'-biphenyl-3-yl)amino]phenyl}ethyl)amino]ethyl}quinolin- 2(1H)-one,

5-[(R )-2-(2-{4-[4-(2-amino-2-methyl-propoxy)-phenylamnino]-phenyl }- ethylamino)-1 -hydroxy-ethyl]-8-hydroxy-1 H-quinolin-2-one ; diuretics, (e.g., amiloride, ; anticholinergics, e.g., ipratropium (e.g., as bromide), tiotropium, atropine or oxitropium); hormones, (e.g., cortisone, hydrocortisone or prednisolone); xanthines, (e.g., aminophylline, choline theophyllinate, lysine theophyllinate or theophylline); therapeutic proteins and peptides, (e.g., insulin). It will be clear to a person skilled in the art that, where appropriate, the medicaments may be used in the form of salts, (e.g., as alkali metal or amine salts or as acid addition salts) or as esters (e.g., lower alkyl esters) or as solvates (e.g., hydrates) to optimise the activity

and/or stability of the medicament. It will be further clear to a person skilled in the art that where appropriate, the medicaments may be used in the form of a pure isomer, for example, R-salbutamol or RR-formoterol.

Particular medicaments for administration using pharmaceutical formulations in accordance with the invention include anti-allergies, bronchodilators, beta agonists (e.g., long-acting beta agonists), and antiinflammatory steroids of use in the treatment of respiratory conditions as defined herein by inhalation therapy, for example cromoglicate (e.g. as the sodium salt), salbutamol (e.g. as the free base or the sulphate salt), salmeterol (e.g. as the xinafoate salt), bitolterol, formoterol (e.g. as the fumarate salt), terbutaline (e.g. as the sulphate salt), reproterol (e.g. as the hydrochloride salt), a beclometasone ester (e.g. the dipropionate), a fluticasone ester (e.g. the propionate), a mometasone ester (e.g., the furoate), budesonide, dexamethasone, flunisolide, triamcinolone, tripredane, (22R)- 6α,9α-difluoro-11β,21-dihydroxy-16α,17α -propylmethylenedioxy-4-pregnen- 3,20-dione. Medicaments useful in erectile dysfunction treatment (e.g., PDE- V inhibitors such as vardenafil hydrochloride, along with alprostadil and sildenafil citrate) may also be employed. It should be understood that the medicaments that may be used in conjunction with the inhaler are not limited to those described herein.

Salmeterol, especially salmeterol xinafoate, salbutamol, fluticasone propionate, beclomethasone dipropionate and physiologically acceptable salts and solvates thereof may be employed.

It will be appreciated by those skilled in the art that the formulations according to the invention may, if desired, contain a combination of two or more medicaments. Formulations containing two active ingredients are known for the treatment and/or prophylaxis of respiratory disorders such as asthma and COPD, and may include, for example, formoterol (e.g. as the fumarate) and budesonide, salmeterol (e.g. as the xinafoate salt) and fluticasone (e.g. as the propionate ester), salbutamol (e.g. as free base or sulphate salt) and beclometasone (as the dipropionate ester).

In one embodiment, a particular combination that may be employed is a combination of a beta agonist (e.g., a long-acting beta agonist) and an antiinflammatory steroid. One embodiment encompasses a combination of

fluticasone propionate and salmeterol, or a salt thereof (particularly the xinafoate salt). The ratio of salmeterol to fluticasone propionate in the formulations according to the present invention is preferably within the range 4:1 to 1:20. The two drugs may be administered in various manners, simultaneously, sequentially, or separately, in the same or different ratios. In various embodiments, each metered dose or actuation of the inhaler will typically contain from 25 μg to 100 μg of salmeterol and from 25 μg to 500 μg of fluticasone propionate. The pharmaceutical formulation may be administered as a formulation according to various occurrences per day. In one embodiment, the pharmaceutical formulation is administered twice daily. Embodiments of specific medicament combinations that may be used in various pharmaceutical formulations are as follows:

1 ) fluticasone propionate 100 μg/ salmeterol xinafoate 72.5 μg (equivalent to salmeterol base 50 μg)

2) fluticasone propionate 250 μg/ salmeterol xinafoate 72.5 μg (equivalent to salmeterol base 50 μg)

3) fluticasone propionate 500 μg/salmeterol xinafoate 72.5 μg (equivalent to salmeterol base 50 μg)

In various embodiments, the pharmaceutical formulations may be present in the form of various inhalable formulations. In one embodiment, the pharmaceutical formulation is present in the form of a dry powder formulation, the formulation of such may be carried out according to known techniques. The invention also encompasses inhalation devices including inhalable formulations. Dry powder formulations for topical delivery to the lung by inhalation may, for example, be presented in capsules and cartridges of for example gelatine, or blisters of for example laminated aluminum foil, for use in an inhaler or insufflator. Powder blend formulations generally contain a powder mix for inhalation of the compound of the invention and a suitable powder base which includes lactose and, optionally, at least one additional excipient (e.g., carrier, diluent, etc.). In various embodiments, each capsule or cartridge may generally contain between 20 μg and 10 mg of the at least one medicament. In one embodiment, the formulation may be formed into

particles comprising at least one medicament, and excipient material(s), such as by co-precipitation or coating. When employed as a dry powder, packaging of the formulation may be suitable for unit dose or multi-dose delivery. In the case of multi-dose delivery, the formulation can be pre- metered (e.g., as in Diskus®, see GB 2242134/ U.S. Patent Nos. 6,032,666, 5,860,419, 5,873,360, 5,590,645, 6,378,519 and 6,536,427 or Diskhaler, see GB 2178965, 2129691 and 2169265, US Pat. Nos. 4,778,054, 4,811 ,731 , 5,035,237) or metered in use (e.g. as in Turbuhaler, see EP 69715, or in the devices described in U.S. Patent No 6,321 ,747). An example of a unit-dose device is Rotahaler (see GB 2064336). In one embodiment, the Diskus® inhalation device comprises an elongate strip formed from a base sheet having a plurality of recesses spaced along its length and a lid sheet hermetically but peelably sealed thereto to define a plurality of containers, each container having therein an inhalable formulation containing the at least one medicament, the lactose, optionally with other excipients. Preferably, the strip is sufficiently flexible to be wound into a roll. The lid sheet and base sheet will preferably have leading end portions which are not sealed to one another and at least one of the leading end portions is constructed to be attached to a winding means. Also, preferably the hermetic seal between the base and lid sheets extends over their whole width. The lid sheet may preferably be peeled from the base sheet in a longitudinal direction from a first end of the base sheet.

In one embodiment, the formulations may be employed in or as suspensions or as aerosols delivered from pressurised packs, with the use of a suitable propellant, e.g. dichlorodifluoromethane, trichlorofluoromethane, dichlorotetrafluoroethane, 1 ,1 ,1 ,2,3,3,3-heptafluoropropane, 1 ,1 ,1 ,2- tetrafluoroethane, carbon dioxide or other suitable gas. Such formulations may be delivered via a pressurized inhaler, e.g., a Metered Dose Inhaler (MDI). Exemplary MDIs typically include canisters suitable for delivering the pharmaceutical formulations. Canisters generally comprise a container capable of withstanding the vapour pressure of the propellant used such as a plastic or plastic-coated glass bottle or preferably a metal can, for example an aluminum can which may optionally be anodised, lacquer-coated and/or

plastic-coated, which container is closed with a metering valve. Aluminum cans which have their inner surfaces coated with a fluorocarbon polymer are particularly preferred. Such polymers can be made of multiples of the following monomeric units: tetrafluoroethylene (PTFE), fluorinated ethylene propylene (FEP), perfluoroalkoxyalkane (PFA), ethylene tetrafluoroethylene (EFTE), vinyldienefluoride (PVDF), and chlorinated ethylene tetrafluoroethylene. Embodiments of coatings used on all or part of the internal surfaces of an MDI are set forth in U.S. Patent Nos. 6,143,277; 6,511 ,653; 6,253,762; 6,532,955; and 6,546,928.

MDIs may also include metering valves are designed to deliver a metered amount of the formulation per actuation and incorporate a gasket to prevent leakage of propellant through the valve. The gasket may comprise any suitable elastomeric material such as for example low density polyethylene, chlorobutyl, black and white butadiene-acrylonitrile rubbers, butyl rubber and neoprene. Suitable valves are commercially available from manufacturers well known in the aerosol industry, for example, from Valois, France (e.g. DF10, DF30, DF60), Bespak pic, UK (e.g. BK300, BK356) and 3M-Neotechnic Ltd, UK (e.g. SpraymiserTM), Embodiments of metering valves are set forth in U.S. Patent Nos. 6,170,717; 6,315,173; and 6,318,603.

In various embodiments, the MDIs may also be used in conjunction with other structures such as, without limitation, overwrap packages for storing and containing the MDIs, including those described in U.S. Patent No. 6,390,291 , as well as dose counter units such as, but not limited to, those described in U.S. Patent Nos. 6,360,739 and 6,431 ,168.

In addition to the above, the pharmaceutical formulations can be employed in capsules, sachets, tablet buccals, lozenges, papers, or other container. Moreover, the formulations can be in the form of tablets, pills, powders, elixirs, suspensions, emulsions, solutions, syrups, capsules (such as, for example, soft and hard gelatin capsules), suppositories, sterile injectable solutions, and sterile packaged powders. Excipients, carriers, diluents, and the like may be optionally employed.

The pharmaceutical formulation formed by the processes of the invention may be used in the treatment of a number of respiratory disorders,

which encompasses, for example, maintenance treatment and/or prophylaxis. Such respiratory conditions include, without limitation, diseases and conditions associated with reversible airways obstruction such as asthma, chronic obstructive pulmonary diseases (COPD) (e.g. chronic and wheezy bronchitis, emphysema), respiratory tract infection and upper respiratory tract disease (e.g. rhinitis, such as allergic and seasonal rhinitis). Such treatment is carried out by delivering medicament to a mammal. Accordingly, and in view of the above, in another aspect, the invention provides a method for the treatment of a respiratory disorder comprising the step of administering a pharmaceutically effective amount of a pharmaceutical formulation to a mammal such as, for example, a human. For the purposes of the invention, the term "pharmaceutically effective amount" is to be broadly interpreted and encompass the treatment of the disorder. In one embodiment, the administration is carried out via an inhalation device described herein. In one embodiment, the administration is carried out by nasal or oral inhalation.

The invention offers potential advantages relative to the prior art. As one example, the invention allows for improved control of particle size distribution of the lactose formed by this process, i.e., a more consistent particle size distribution and/or a more consistent surface morphology of the lactose is capable of being achieved from lactose feeds independent of the particle size distribution of the lactose feed. In particular, the lactose formed by the process of the invention is capable of exhibiting a more continuous particle size distribution, i.e., very little or no gap in the particle size distribution in contrast to the distribution set forth in X. M. Zeng et al., International Journal of Pharmaceutics, 176 (1998) 99-110.

The present invention is highly advantageous. In one embodiment, for example, the classification cut-point can be chosen so that the coarse fraction is suitable for use in a pharmaceutical formulation without fine fraction. In one embodiment, at least a portion of the fine fraction (ranging from 0, 5, 10, 15, 20, 25, 30, 35, 40, or 45 to 55, 60, 65, 70, 75, 80, 85, 90, 95 or 100 percent by weight) may be combined (e.g., blended) with at least a portion of the coarse fraction (ranging from 0, 5, 10, 15, 20, 25, 30, 35, 40, or 45 to 55, 60, 65, 70, 75, 80, 85, 90, 95 or 100 percent by weight) such that the resulting lactose composition is suitable for use in a pharmaceutical formulation. In one

embodiment, the fine fraction may be used in a pharmaceutical formulation without any coarse fraction. With respect to the above embodiments, it is preferred that lactose be employed with sufficient fine material so as to be capable of meeting the FPMass requirements, typically 20 percent to 30 percent (weight basis) of the medicament content achieved by utilising lactose that comprises 2 percent to 10 percent (volume basis) particles less than 4.5 microns as measured by Sympatec.

In view of the above, in one embodiment, the invention may further comprise combining at least one medicament with a lactose composition to form a pharmaceutical formulation. The lactose composition in such an embodiment may comprise from 0, 5, 10, 15, 20, 25, 30, 35, 40, or 45 to 55, 60, 65, 70, 75, 80, 85, 90, 95 or 100 percent by weight of the coarse fraction and from 0, 5, 10, 15, 20, 25, 30, 35, 40, or 45 to 55, 60, 65, 70, 75, 80, 85, 90, 95 or 100 percent by weight of the fine fraction. Moreover, in one embodiment, the invention may further comprise combining at least a portion of the coarse fraction with at least a portion of the fine fraction to form a lactose composition; and thereafter combining the lactose composition with at least one medicament to form a pharmaceutical formulation. Alternatively, in one embodiment, the invention may encompass simultaneously combining: (i) from 0, 5, 10, 15, 20, 25, 30, 35, 40, or 45 to 55, 60, 65, 70, 75, 80, 85, 90, 95 or 100 percent by weight of coarse fraction, (ii) from 0, 5, 10, 15, 20, 25, 30, 35, 40, or 45 to 55, 60, 65, 70, 75, 80, 85, 90, 95 or 100 percent by weight of fine fraction, and (iii) at least one medicament to form a pharmaceutical formulation.

As the large majority of the small particles have been produced by milling, they all typically possess a similar surface roughness. By comparison, if lactose batches are introduced into the mill that contain particles less than a volume average size of 70 microns, then these particles will essentially pass through the mill without the need to be reduced in size by the milling action. As a consequence, these particles will still exhibit a smoother crystalline surface, compared to milled lactose particles, which in turn may lead to different interactions with the active molecules.

Thus by substantially reducing or eliminating these small crystalline particles from the feed to the mill, not only can the milled particle size be

controlled, but the particles will also exhibit a very much more uniform surface morphology.

Malvern Methodology

Malvern measurements as set forth herein are determined according to the following:

Equipment

Malvern Mastersizer, laser diffraction particle size analyser fitted with the 300mm focal length lens.

MS15 stainless steel flow-through measurement cell, 2.2 mm path width (Malvern).

MSX1 small volume presentation unit (Malvern).

Viton tubing: 4.8 mm internal diameter, 1.6 mm wall thickness (Watson-

Marlow).

Viton or Neoprene tubing: approximately 15mm internal diameter, 2 mm wall thickness (Watson-Marlow).

Positive pressure or vacuum filtration equipment with 0.22 μm maximum aperture membrane filters type GV (Sartorius/Millipore).

Analytical balance (Mettler, electronic) minimum weighing capacity of

30 grams with an accuracy to 4 decimal places,

Spatulas for sample transfer.

Solvent : HPLC grade iso-octane

Test Method

This section outlines the procedure for the measurement of one batch of lactose. For each batch, it is preferred that three representative samples be measured in duplicate.

Preparation of the liquid dispersant.

For the preparation of 2 litres of liquid dispersant, dissolve with ultrasonication 0.7 g of lecithin NF in 10cm ³ of iso-octane to produce a dispersant concentrate.

Filter sufficient iso-octane using the vacuum filtration unit and a 0.22 μm filter disc. During this filtration add the dispersant concentrate

to the solvent so that it too is filtered and mixed in. Store in a pre- cleaned Winchester solvent container (Liquid dispersant).

The filtered liquid dispersant should be stored in a solvent cupboard when not in use.

Liquid dispersant may be recycled by refiltering after analysis as from above and making up to volume with iso-octane

Preparation of the Malvern Mastersizer

Set up the following measurement parameters on the Malvern Mastersizer for the test sample :

Set up the MSX1 small volume presentation unit ensuring that the total length Viton tubing from the presentation unit to the flow through cell does not exceed 30cm. Attach one end of suitably sized (int. dia. approx. 15mm) Viton or Neoprene tubing (waste tubing) to the waste drain and place the other end into a red non-chlorinated waste solvent container placed on the floor.

Flush the presentation unit and the cell through with the liquid dispersant and drain.

Refill with the iso-octane/lecithin dispersant and align the laser beam using a stirrer speed of 11 o ¹ clock (1500 ± 100 rpm). The laser intensity should be in the "good" region.

Once the laser is aligned, it is preferred that the stirrer speed should not be altered.

Sample measurement

Add the powder sample to a presentation unit using a spatula to obtain an obscuration value within the range 0.18 to 0.22.

Allow 60 seconds to allow the dispersion to equilibrate and the obscuration value to stabilise.

Measure the particle size distribution. Repeat measurement as deemed appropriate.

Svmpatec Methodology

The following teachings may be employed to arrive at the Sympatec measurements described herein: Instrumental Parameters

The Sympatec HELOS is provided with the RODOS dry powder dispersion unit and the VIBRI vibratory feeder.

In these measurements, software is used in conjunction with the measurement equipment.

For the RODOS dispersing system, the primary pressure of the injector should be adjusted using the pressure control dial. The primary pressure should be within the range 1.3 - 1.7 bar although a pressure of 1.5 bar should be aimed for at each run. The injector depression should be optimised using the adjustment ring. The direction in which the adjuster ring is turned (clockwise or anti clockwise), has no adverse effect on the depression obtained. On instruments with a RODOS/M dispersing system, the primary pressure may be adjusted using a software algorithm. The injector depression should be maximised by clicking the "Auto-adjust depr" button. The instrument should not be used if the injector depression is less than 55mbar at 1.3 - 1.7 bar.

Analysis of sample

Homogenize the sample by turning the pot slowly end-over-end at least ten times in one direction and ten times in the other direction prior to analysis. This is only necessary the first time a sample is taken from a pot.

Transfer 2 ± 1 g of the lactose sample into the funnel attached to the VIBRI using a Kartell "spoon/flat" general purpose spatula (Fisher catalogue no SMJ-410-091 M, volume approximately 1.8 cm ³ ).

A heaped spoonful of lactose powder has been demonstrated to present a sample in the range 2.0 - 3.Og.

The invention will now be described with respect to the following examples. It should be appreciated that the examples are set forth for illustrative purposes only, and do not limit the scope of the invention as defined by the claims.

Example 1 Lactose Processing

Two batches of coarse crystalline lactose A were selected. These differed from each other inasmuch that they exhibited different D50's. Batch 1 had a D50 of between 110 - 130 microns and Batch 2 had a D50 of between 160 - 180 microns, as measured by Malvern laser particle sizing techniques.

Each of these batches were then separated into a fine and coarse fraction using an air classifier. The cut point was nominally about 80 microns.

In each case the fine fraction was discarded and the coarse fractions were milled in an air classifier mill (ACM) so that the D50 of the milled product was nominally 60 - 70 microns.

The milled lactose batches were then classified at a cut point of nominally 25 microns to form a fine fraction and a coarse fraction.

Suitable proportions of the coarse and fine fraction from each milled batch were blended together so that the percentage of lactose < 4.5 microns was:

(i) 6.7% +/- 0.3%; ex batch 1 (ii) 5.0% +/- 0.3%; ex batch 1 (iii) 3.3% +/- 0.3%; ex batch 1

(iv) 6.7% +/- 0.3%; ex batch 2(v) 5.0% +/- 0.3%; ex batch 2 (vi) 3.3% +/- 0.3%; ex batch 2

The above teachings are exemplified by the following protocol:

The purpose of the following development protocol is to manufacture lactose batches having the same overall particle size distribution from different feed crystalline lactose A batches.

The feed, milling, classifying, and blending of lactose may be carried out using the following:

1) A pilot scale blender capable of blending 100 kg of material ("pilot scale blender")

2) A full scale classifier, e.g., a classifier that may be employed as a stand alone device suitable for full scale production ("full scale classifier")

The following experimental method may be employed: 1. Select 2 batches of crystalline lactose A that are different from each other in their percentage of lactose that is less than 80 microns.

Crystalline Lactose A is employed as feed lactose and is made commercially available by Borculo Domo Ingredients of Borculo, Netherlands.

These are defined as batch 1 , which has a D50 of between 120 - 130 microns and batch 2 that has a D50 of 170 - 180 microns. Batch 1 should have the percentage less than 15 microns greater than 3%, whereas batch 2 should have the percentage of lactose less than 15 microns as less than 1.5%.

2. Using the full scale classifier, separate each full batch as cleanly as possible into two fractions, one being < 80 microns and the other being > 80 microns. The same settings should be used for both batches and it is desirable that no changes are to be made to the settings during the classification.

3. Mill only the fraction of lactose that is > 80 microns from both batches 1 and 2. Set the mill at nominally the same as used for commercial supply however, reduce the throughput rate of the lactose so that the milling and built in classifier can act more efficiently. Once set, discard any test millings and then mill each batch of lactose> 80 microns at the same settings.

4. Using the full scale classifier, separate each milled batch as cleanly as possible into two fractions, one being nominally < 25 microns and the other being nominally > 25 microns. The same settings should be used for both batches and it is desirable no changes are to be made to the settings during the classification.

5. Recombination by blending: Using appropriate aliquots from each batch of the classified fractions, recombine the lactose in the pilot scale blender to give:

a - IOOkgs of product which has the percentage of lactose < 4.5 microns at 6.7% +/-0.3% - From feed batch 1 b - IOOkgs of product which has the percentage of lactose < 4.5 microns at 5.0% +/-0.3%- From feed batch 1

c - IOOkgs of product which has the percentage of lactose < 4.5 microns at 3.3% +/-0.3%- From feed batch 1

This may be repeated using lactose from the second feed batch, (d) should be as similar to (a) as possible.

d - IOOkgs of product which has the percentage of lactose < 4.5 microns at 6.7% +/-0.3%- From feed batch 2

e - IOOkgs of product which has the percentage of lactose < 4.5 microns at 5.0% +/-0.3%- From feed batch 2 f - IOOkgs of product which has the percentage of lactose < 4.5 microns at 3.3% +/-0.3%- From feed batch 2

6. Dispense lactose from blender into 20kg tie sealed polyethylene bags, heat sealed in a foil laminate bag and pack inside a cardboard box. Pack at least 5 boxes from each batch.

Example 2 Lactose Processing

Two batches of coarse crystalline lactose A were selected. These differed from each other inasmuch that they exhibited different D50's. Batch 1 had a D50 of between 110 - 130 microns and Batch 2 had a D50 of between 160 - 180 microns, as measured by Malvern laser particle sizing techniques.

Each of these batches were then separated into a fine and coarse fraction using an air classifier. The cut point was nominally about 80 microns.

In each case the fine fraction was discarded and the coarse fractions were milled in an air classifier mill (ACM) so that the D50 of the milled product was nominally 60 - 70 microns.

Both milled lactose batches were then put through a 150 micron sieve. This is an additional step compared to Example 1.

The milled lactose batches were then classified at a cut point of nominally 25 microns to form a fine fraction and a coarse fraction.

Suitable proportions of the coarse and fine fraction from each milled batch were blended together so that the percentage of lactose < 4.5 microns was:

(i) 6.7% +/- 0.3%; ex batch 1 (ii) 5.0% +/- 0.3%; ex batch 1 (iii) 3.3% +/- 0.3%; ex batch 1 (i) 6.7% +/- 0.3%; ex batch 2 (ii) 5.0% +/- 0.3%; ex batch 2 (iii) 3.3% +/- 0.3%; ex batch 2

The unit operation of this protocol pertains to the process of classification of feed, milling, classifying, sieving, and blending of lactose.

The following equipment may be used: pilot scale blender full scale classifier

The purpose of this example is to manufacture lactose batches with the same overall particle size distribution from different feed batches. This work will use the Full Scale Classifier.

The above teachings are exemplified by the following protocol:

1. Select two batches of crystalline lactose A, batch 1 that has a D50 of 120 -130 microns, and the percentage lactose less than 15 microns is greater than 3%, and batch 2 with an D50 of 170 - 180 microns and the percentage lactose less than 15 microns is less than 1.5%. For the purposes of this protocol however, the D50 range was altered such that the batch 1 D50 was from 110-130 microns and the batch 2 D50 was from 160-180 microns.

Treat each of these big bags in an identical fashion as described in stages 2 through to 5.

2. Using the full scale classifier, separate the whole batch as cleanly as possible into two fractions, one being < 80 microns and the other being > 80 microns. The same settings should be used as those used in Example 1 and it is desirable that no changes be made to the settings during the classification.

3. Mill the coarse fraction obtained from 2. Set the mill at nominally the same as used for typical lactose milling. Once the mill is set, discard any test millings and then mill the coarse fractions from the batch at the same settings.

4. Sieve the milled material through a 150-micron sieve.

5. Using the full scale classifier, separate each sieved milled batch as cleanly as possible into two fractions, one being nominally < 25 microns and the other being nominally > 25microns. These settings should remain constant during the classification of the batch.

6. Recombination bv blending: Using appropriate coarse and fine aliquots from batches 1 and 2 at the end of stage 5, recombine the lactose in the pilot scale blender to give: a - IOOkgs of product which has the percentage of lactose < 4.5 microns at 6.7% +/-0.3% - (batch 1 , Fine Feed) b - IOOkgs of product which has the percentage of lactose < 4.5 microns at 6.7% +/-0.3%- (batch 2, Coarse Feed)

No Targets for the D50 are given for these recombinations, 60 - 90 microns would be ideal.

7. Dispense lactose from blender into 20kg tie sealed polythene bags, heat sealed in a foil laminate bag and pack inside a cardboard box. Pack at least 5 boxes from each batch.

Example 3 Lactose Processing

Two batches of coarse crystalline lactose A were selected. These differed from each other inasmuch that they exhibited different D50's. Batch 1 had a D50 of between 110 - 130 microns and Batch 2 had a D50 of between 160 - 180 microns, as measured by Malvern laser particle sizing techniques.

Each of these batches were then separated into a fine and coarse fraction using an air classifier. The cut point was nominally about 80 microns.

In each case the fine fraction was discarded and the coarse fractions were milled in an air classifier mill (ACM) so that the D50 of the milled product was nominally 60 - 70 microns.

The milled lactose batches were then classified at a cut point of nominally 25 microns to form a fine fraction and a coarse fraction.

Suitable proportions of the coarse and fine fraction from each milled batch were blended together so that the percentage of lactose < 4.5 microns was:

(i) 6.3% +/- 0.3%; ex batch 1 (ii) 6.0% +/- 0.3%; ex batch 1 (Hi) 5.5% +/- 0.3%; ex batch 1 (i) 6.3% +/- 0.3%; ex batch 2 (ii) 6.0% +/- 0.3%; ex batch 2 (iii) 5.5% +/- 0.3%; ex batch 2

The above teachings are exemplified by the following protocol:

The unit operation of this protocol pertains to the process of classification of feed, milling, classifying, sieving, and blending of lactose.

The purpose of the protocol set forth in this example is to manufacture lactose batches with the same overall particle size distribution from different feed batches. This work will use the full scale classifier.

The protocol is as follows:

1. Select two batches of crystalline lactose A, batch 1 that has a D50 of 120 - 130 microns, and the percentage lactose less than 15 microns is greater than 3%, and batch 2 with an D50 of 170 - 180 microns and the percentage lactose less than 15 microns is less than 1.5%. For the purposes of this protocol however, the D50 range was altered such that the batch 1 D50 was from 110-130 microns and the batch 2 D50 was from 160-180 microns.

Treat each of these big bags in an identical fashion as described in stages 2 through to 4.

2. Using the full scale classifier, separate the whole batch as cleanly as possible into two fractions, one being < 80 microns and the other being > 80 microns. The same settings should be used as those used in Example 2, these are shown in Table 3 below. It is preferable that no changes are to be made to the settings during the classification.

Table 3

Settings Readings Calculated

Classifierspeed 4.1 Hz 246 rpm

Feedrate 15 Hz 12 Rpm 250 Kg/hour

Transport air 1500 Rpm 25 Hz min 135m3/h max 165 m3/h

Washair 1500 Rpm 25 Hz min 390m3/h max 450 m3/h

3. Using the full scale classifier with the settings in Table 4, separate each milled batch as cleanly as possible into two fractions, one being nominally < 25 microns and the other being nominally > 25microns.

Table 4

Settings Readings Calculated

Classifierspeed 10 Hz 600 rpm

Feedrate 25 Hz 20 Rpm 410 KG/hour

Transport air min 175m3/h max 235 m3/h

Washair min 620m3/h max 675 m3/h

Airsystem 2100 Rpm 35 Hz

The above settings are capable of providing batches that are approximately 80% wt/wt coarse material and 20% wt/wt fine material.

4. Recombination by blending: Using appropriate coarse and fine aliquots from batches 1 and 2 at the end of stage 3, recombine the lactose in the pilot scale blender to give:

a - IOOkgs of product which has the percentage of lactose < 4.5 microns at 6.3% +/-0.3% - From feed batch 1 b - IOOkgs of product which has the percentage of lactose < 4.5 microns at 6.0% +/-0.3%- From feed batch 1 c - IOOkgs of product which has the percentage of lactose < 4.5 microns at 5.5% +/-0.3%- From feed batch 1

This to be repeated using lactose from the second feed batch.

d - IOOkgs of product which has the percentage of lactose < 4.5 microns at 6.3% +/-0.3%- From feed batch 2 e - IOOkgs of product which has the percentage of lactose < 4.5 microns at 6.0% +/-0.3%- From feed batch 2 f - IOOkgs of product which has the percentage of lactose < 4.5 microns at 5.5% +/-0.3%- From feed batch 2

- Use the same amounts of fines and coarse in each pair of the combinations.

- Ensure data on Malvem and Sympatec sheets agree with data in the table.

Table 5 - Particle Size Data for Typical Lactose Production

An important aspect of the parameters described within the protocols illustrated in Examples 1 through 3 is that all initial stages are the same, irrespective of the desired amount of fine lactose in the final product to be manufactured. The varying amounts of fine lactose that are desired are achieved by combining the appropriate proportions of the coarse and fines fractions and not to mill "harder" for the lactose batches that require more fine lactose in the final product. Tables 6 and 7 show the particle size for the feed Batch A crystal by both Malvern and Sympatec, that was used in these trials.

It can be seen from these three examples that all the initial stages are similar, if not identical. The only stage at which there is a change is at the blending stage, in which more fine lactose would be added to achieve the required targets at %< 4.5 microns.

Tables 6 and 7 show the particle size data for the both coarse and fine feed crystal lactose used in Examples 1 , 2 and 3. (Malvern and Sympatec)

Table 6 - Particle size data of the feed lactose prior to classification -

Malvern

Discussion:

A range of feed lactose has been covered by choosing both coarse and fine feed lactose in each of these three examples. They contain varying amounts of lactose that would pass through the mill, essentially un-milled, and end up in packed product.

In each of the examples 1-3, the majority of these small particles are removed by a classification process so that the resultant feed material is much more uniform when presented to the mill. A target was to reduce the amount of lactose particles less than 70 microns to less than 10%. An alternative way of expressing this is that the D10 of the resulting lactose should be greater than 70 microns by Sympatec.

The results from the three examples are shown in Table 8.

Table 8 - Coarse Fraction from the Crystalline Feed Lactose after initial classification step - Sympatec

It can be seen from Table 8 that there were two instances in which the target for removing the fine material less than 70 microns was not quite achieved, although the values of 66 and 67 microns are close enough that they are believed to not impact on the outcome of the experiments.

Milling of the Classified Feed

Each of the coarse fractions were then milled to reduce the D50 to between 60 - 70 microns by Malvern. The values from Example 1 indicated that slightly harder milling was required, so this was adjusted for Examples 2 and 3.

The results are shown in Table 9 (Malvern) and Table 10 (Sympatec).

Table 9 - Milled "Coarse" fraction. Particle Size Da bv Malvern.

Classification of the Milled Lactose

The milled lactose was then separated into a coarse and fine fraction using the same classifier as was used to separate the fine lactose from the input feed lactose. The resultant particle size data are shown in Tables 11 , 12, 13 and 14.

Table 11 - Fine Fractions after Milling and Classification of the Coarse Crystalline Feed Fraction by Malvern

Table 12 -Coarse Fractions after Milling and Classification of the Coarse Crystalline Feed Fraction by Malvern

Table 13 -Fine Fractions after Milling and Classification of the Coarse Crystalline Feed Fraction by Sympatec

Table 14 - Coarse Fractions after Milling and Classification of the Coarse Crystalline Feed Fraction by Svmpatec

Recombination by Blending

The Examples 1 - 3 then call for appropriate portions of the fine and coarse fractions to be blended together so that the resultant blend contains the required amount of lactose that is less than 4.5 microns. This is in order to achieve the desired value for the FPMass performance for an inhalable formulation of salmeterol xinafoate and fluticasone propionate. These target values are listed in the Examples 1-3, and the results are shown in Tables 15-17.

Table 15- Particle Size Data for final packed product from Example 1-

Sympatec

It can be seen from these Example 1 data that the Sympatec targets at %< 4.5 microns were met in every case.

Table 16 -Particle Size Data for final packed product from Example 2 -

Svmpatec

It can be seen from the Example 2 particle size data that the Sympatec targets were met in every case.

Table 17 -Particle Size Data for final packed product from Example 3-

Svmpatec

It can be seen from these Example 3 data that the Sympatec targets were met for the three batches stemming from the coarse feed but not for those made from the fine feed.

Averages and relative standard deviation values for the D10, D50, D90 and the %< 15 microns for both the fine and coarse fractions for conventional production were evaluated, compared to the process of the invention . These are shown in Table 17. All values are Malvern data and the normal production is all batches of the same lactose product manufactured over an 18 month period.

Table 17 - Assessment of Particle size consistency for Example 1.

The results in Table 17 show that the particle size of the coarse and fine fractions are more consistent than that obtained during normal production, despite the fact that an adjustment was made to the milling settings between Example 1 and Examples 2 and 3. If Example 1 data is eliminated from the assessment then the variation in particle size is even less as shown in Table 18.

Table 18 - Assessment of Example 2 and Example 3 data only-Malvern

Particle Size Profiles for Example 1. Example 2 and Example 3

The full profiles for each of the three pairs of lactose made in Examples 1 and 3 are shown in FIGS. 1-3.

It can be seen that these three pairs of curves are reasonably similar, despite originating from a "coarse" and "fine" feed in each occasion. For a comparison FIG. 4 shows the data form Example 2 overlaid with two normal production batches of similar lactose. These show the extremes in differing profiles that can be obtained during normal production.

Example 4 Lactose Production

The following example demonstrates the production of lactose suitable for use in inhalation formulations, which has been produced by a method that does not require the combination of any of the fine fraction to the coarse fraction.

The following experimental method may be employed: 1. Classify 2 tons of crystalline lactose A so that the majority of fine crystals < 70 microns are removed into the fine fraction.

The remaining coarse fraction must comply with the acceptance criteria shown in Table 19 below.

Table 19

Measurement are by Sympatec %<4.5 microns D10 D50

Coarse Fraction <0.2% > 70 microns >130 microns

3. Mill only the coarse fraction of lactose targetting 60 microns +/- 3 microns for the D50 of the milled product.

4. Using pilot scale blender, separate 300kgs of the milled batch into a fine and coarse fraction using the maximum classifier wheel speed and suitable air flow parameters to achieve the lowest micron cut point that can be achieved. This will leave a coarse fraction that contains as much fine lactose as possible. (Product A)

5. Repeat 4 on another 300kg aliquot of the milled lactose, but reduce the speed of the classifier wheel and keep the air flow at the same settings. This will produce a somewhat coarser coarse fraction. (Product B)

6. Repeat 4 again on another 300kg aliquot of the milled lactose, but reduce the speed of the classifier wheel further and keep the air flow at the same settings. This again will produce an even coarser coarse fraction. (Product C)

7. Do not add any of the fine fractions to the coarse fractions, but blend each of the coarse fractions to ensure homogeneity. Dispense the blended lactose into 20kg tie sealed polyethylene bags, which are then heat sealed into a foil laminate bag.

Pack at least 7 x 20kgs boxes of each batch, and record the particle size distribution by Sympatec on the final product.

Table 20 - Results of the Classified Feed Lactose - Coarse Fraction:

It can be seen from the data in Table 20 that the readings comply with the specifications detailed in Table 19.

Results for the full particle size distribution of the final products A, B & C by Sympatec are in Table 21.

Table 21

These data can also be presented graphically as shown in Figure 8. It can be seen from Figure 8 that the amount of lactose present in the region < 25 microns high size is of the same order as that shown in Figure 4 which is that of routine commercial batches used for inhalation grade lactose. Thus, these coarse fractions are believed to be capable of displaying suitable Fine Particle Mass (FPM) performance to make them useful as inhalation grade lactose.

Conclusions:

All of the trials, Example 1 , Example 2 and Example 3 have been assessed for Cl performance at release and upon accelerated ageing at 25 degrees Celsius, 75% relative humidity. The series of Tables 22A, 22B and 22C show the initial mean data that were obtained during the stability study for Fine Particle Mass (FPMass), Throat, Pre-separator and stage 0, (TPO) and Stages 3+4.

Table 22A i Initial Mean Fine Particle Mass Data for 75 μg salmeterol xinafoate (equivalent to 50 μg salmeterol base) and 100 μg fluticasone propionate formulation using a conventional fine lactose

Table 22A ii Initial Mean TPO Data for 50μg salmeterol base/100 μg fluticasone propionate formulation sing a conventional fine lactose

NB. Specifications for Advair 50/100 are:

FP FPMass, 21 - 30; SX FPMass, 9 - 13

FP Stages 3 + 4, 11 - 19; SX Stages 3 + 4, 4 - 8

FP TPO, 55 - 80; SX TPO, 28 - 42

Table 22B i Initial Mean Fine Particle Mass Data for 50μg salmeterol base/250 μg fluticasone propionate formulation using a conventional medium grade lactose

Table 22B ii Initial Mean TPO Data for 50μg salmeterol base/250 μg fluticasone propionate formulation using a medium grade lactose

Table 22B iii Initial Mean Stages 3+4 Data for 50μg salmeterol base/250 μg fluticasone propionate formulation using a conventional medium grade lactose

NB. Specifications for Advair 50/250 are:

FP FPMass, 51 - 75; SX FPMass, 9 - 13

FP Stages 3 + 4, 29 - 48; SX Stages 3 + 4, 5 - 8

FP TPO, 140 - 200; SX TPO, 28 - 42

Table 22C i Initial Mean Fine Particle Mass Data for 50μg salmeterol base/500 μg fluticasone propionate formulation using a conventional coarse grade lactose

Table 22C ii Initial Mean TPO Data for 50μg salmeterol base/500 μg fluticasone propionate formulation using a conventional coarse grade lactose

Table 22C iii Initial Mean Stages 3+4 Data for 50μg salmeterol base/500 μg fluticasone propionate formulation using a conventional coarse grade lactose

NB. Specifications for Advair 50/500 are:

FP FPMass, 106 - 150; SX FPMass, 10 - 14.

FP Stages 3 + 4, 60 - 95; SX Stages 3 + 4, 5 - 9. FP TPO, 290- 400; SX TPO, 28 - 42.

Taking all of the factors into account, the data suggests that the pre- classification of the feed crystal to remove small lactose crystals below approximately 70 microns does not appear to significantly alter the performance ex device when compared to the normal method of lactose production. Additionally, and advantageously, the invention offers a process for forming lactose which provides improved control of particle size distribution relative to conventional processes.

Example 5 Stability Tests

Stability data for various formulations each containing 100 μg fluticasone propionate and 50 μg salmeterol base along with lactose formed in accordance with the invention were evaluated under naked conditions at 25°C and 75 percent Relative Humidity. As shown in Table 23, such data were evaluated at different time points. As shown, formulations made from the process according to the invention exhibit good stability characteristics.

Table 23

Examples 6-8 Stability Studies

Examples 6-8 set forth various pharmaceutical formulations for which the stability was evaluated. Batches 01 , 02 and 03 were processed in accordance with the present invention by employing a pre-classification stage prior to proceeding by conventional manufacturing used for making Advair® made commercially available by GlaxoSmithkline of Brentford, United Kingdom. Batch 04 was made using an Advair® manufacturing process. In Tables 24-28, the symbol "T" indicates that the particle size distribution of salmeterol xinafoate and fluticasone propionate were evaluated by Cascade Impaction according to Method STM-195.

Tables 29-31 illustrate the average change of drug recovered in stages 1 to 5 as a percentage, normalising the results to take into account the total drug recovery for each device. The change is determined between the initial and final time points as defined in the protocol and is calculated according to the following equation:

Stage 1 to 5 recovery at initial x 100 - Stage 1 to 5 recovery at final time point x 100

Total recovery at initial Total recovery at final time point

Tables 32-34 provide summaries of the stability studies that are set forth in Tables 35 to 130. These tables illustrate Cascade Impaction results.

Referring to the tables that follow:

"500/50 μg" refers to a formulation containing 50 μg salmeterol base and 500 μg fluticasone propionate in lactose.

"250/50 μg" refers to a formulation containing 50 μg salmeterol base and 250 μg fluticasone propionate in lactose.

"100/50 μg" refers to a formulation containing 50 μg salmeterol base and 100 μg fluticasone propionate in lactose.

"MDPI" refers to a multidose dry powder inhaler.

A* refers to in-use testing. Devices to be removed from 25°C/60% RH storage and overwrappers to be removed. Samples to be returned to storage at 25°C/75% RH. See Tables 25 and 27.

In Tables 32-34, Batches A, E and I correspond to batch 01. Batches B, F and J correspond to batch 02. Batches C, G and K correspond to batch 03. Batches D, H and L correspond to batch 04.

Tables 24-25: Testing Protocols for 1 UU/bU μg

Table 24

Table 25

Table 26: Testing Protocol for 250/50 μg

Tables 27-28 Testing Protocol for 500/50 μg

Table 27

Table 28

O

Table 29: Average Change of Drug Recovered in Stages 1-5 (100/50 μg )

Table 30: Average Change of Drug Recovered in Stages 1-5 (250/50 μg )

Table 31 : Average Change of Drug Recovered in Stages 1-5 (500/50 μg )

K>

Table 32 Summary of Stability Studies for 100/50μg MDPI (Example 6)

Ul

Table 33 Summary of Stability Studies for 250/50μg MDPI (Example 7)

CΛ

Table 34 Summary of Stability Studies for 500/50μg MDPI (Example 8)

Ul Ul

Table 35 Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: A Storage condition

25°C/60%RH

Testing Complete

Table 36 Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number; A Storage condition 25°C/60%RH

Testing Complete

End of life Storage Time Specification Initial 9 months 12 months 15 months

-4

(Contd)

Table 37. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: A Storage condition 25°C/60%RH

Testing Complete

OO

(Contd)

Table 38. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: Storage condition

25°C/60%RH

Testing Complete

Ul

Table 39. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap

Packed Batch Number. A Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

o

Note: NGT = Not Greater Than {Contd)

Table 40. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap

Packed Batch Number: Storage condition

25°C/75%RH

Testing Complete

c \

Table 41. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 14 months

Packed Batch Number: A Storage condition

25°C/75%RH

Testing Complete

End of life Storage Time Specification Initial 2 weeks 1 month 1.5 months

κ>

(Cont'd)

Table 42. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 14 months (cont'd)

Packed Batch Number: A Storage condition

25°C/75%RH

Testing Complete

Table 43. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number A Storage condition

25°C/60%RH

Testing Complete

Test End of life Storage Time

(Contd)

Table 44. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: B Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time

Specification Initial I 9 months | 12 months 15 months

Ul

(Contd)

Table 45. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: B Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time

(Contd)

Table 46. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: B Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time Specification Initial 9 months 12 months 15 months

as

-4

Note: NGT = Not Greater Than

Table 47. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap

Packed Batch Number: B Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

Specification Initial 2 Weeks 1 Month 1.5 Months | 3 Months

OO

(Contd)

Table 48. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap

Packed Batch Number: B Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time Specification Initial 2 Weeks 1 Month 1.5 Months 3 Months

Os

Table 49. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 14 months

Packed Batch Number: B Storage condition

25°C/75%RH

Testing Complete

-4

O

(Contd)

Table 50. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 14 months

Packed Batch Number: B Storage condition

25°C/75%RH

Testing Complete

Table 51. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: C Storage condition 25°C/60%RH

Testing Complete

-4 K>

(Contd)

Table 52. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: C Storage condition 25°C/60%RH

Testing Complete

-4

(Coπtd)

Table 53. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: C Storage condition 25°C/60%RH

Testing Complete

(Contd)

Table 54. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: C Storage condition 25°C/60%RH

Testing Complete

-4

Table 55. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap

Packed Batch Number: C Storage condition

25°C/75%RH

Testing Complete

-4

Table 56. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap

Packed Batch Number: C Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

-4 -4

Table 57. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 14 months

Packed Batch Number: C Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

-4

(Contd)

Table 58. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 14 months

Packed Batch Number: C Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

-4

Table 59. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: D Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time Specification Initial 1 Month 3 Months 6 Months

O

Note: NGT = Not Greater Than (Coπtd)

Table 60. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: D Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time

OC

(Contd)

Table 61. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: D Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time Specification Initial 1 Month 3 Months 6 Months

so κ>

(Contd)

Table 62. Stability data for DISKUS® 100/50 μg 60 Dose Product with overwrap

Packed Batch Number: D Storage condition 25°C/60%RH

Testing Complete

Table 63. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap

Packed Batch Number: D Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

Specification Initial 2 Weeks I 1 Month I 1.5 Months | 3 Months

OO

(Contd)

Table 64. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap

Packed Batch Number: D Storage condition

25°C/75%RH

Testing Complete

OO Ul

Table 65. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 14 months

Packed Batch Number: D Storage condition

25°C/75%RH

Testing Complete

(Contd)

Table 66. Stability data for DISKUS® 100/50 μg 60 Dose Product without overwrap (Cont'd) Samples previously stored at 25°C/60%RH with overwrap for 14 months

Packed Batch Number: D Storage condition

25°C/75%RH

Testing Complete

OO -4

Table 67. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: E Storage condition

25°C/60%RH

Testing Complete

Test End of life Storage Time

OO 90

(Coπtd)

Table 68. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: E Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time

90

(Contd)

Table 69. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: E Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time Specification Initial 1 month 3 months 6 months

O

(Contd)

Table 70. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: E Storage condition

25°C/60%RH

Testing Complete

Test End of life Storage Time Specification initial 9 months 12 months 18 months

Table 71. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Packed Batch Number: E Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

(Contd)

Table 72. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Packed Batch Number: Storage condition

25°C/75%RH

Testing Complete

Table 73. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH for 17 months with overwrap

Packed Batch Number: Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

(Contd)

Table 74. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Samples previously stored at 25°C/60%RH for 17 months with overwrap (cont'd)

Packed Batch Number: Storage condition

25°C/75%RH

Testing Complete

Table 75. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: F Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time

(Contd)

Table 76. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: F Storage condition

25°C/60%RH

Testing Complete

Test End of life Storage Time

-4

(Contd)

Table 77. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: F Storage condition

25°C/60%RH

Testing Complete

OO

(Contd)

Table 78. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: F Storage condition

25°C/60%RH

Testing Complete

Test End of life Storage Time

Table 79. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Packed Batch Number: Storage condition

25°C/75%RH

Testing Complete

O O

(Con(d)

Table 80. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Packed Batch Number: F Storage condition

25°C/75%RH

Testing Complete

Table 81. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH for 17 months with overwrap

Packed Batch Number: F Storage condition

25°C/75%RH

Testing Complete

O κ>

(Contd)

Table 82. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Samples previously stored at 25°C/60%RH for 17 months with overwrap (cont'd)

Packed Batch Number: F Storage condition

25°C/75%RH

Testing Complete

O

Table 83. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time

O 4-

(Contd)

Table 84. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: G Storage condition

25°C/60%RH

Testing Complete

O Ul

(Contd)

Table 85. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: G Storage condition 25°C/60%RH

Testing Complete

O

(Contd)

Table 86. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: G Storage condition

25°C/60%RH

Testing Complete

O o

-4

Table 87. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Packed Batch Number: G Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

O SC

(Contd)

Table 88. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Packed Batch Number: Storage condition

25°C/75%RH

Testing Complete

O

Table 89. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH for 17 months with overwrap

Packed Batch Number: G Storage condition

25°C/7S%RH

Testing Complete

(Contd)

Table 90. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Samples previously stored at 25°C/60%RH for 17 months with overwrap (cont'd)

Packed Batch Number: G Storage condition

25°C/75%RH

Testing Complete

Table 91. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: H Storage condition

25°C/60%RH

Testing Complete

(Contd)

Table 92. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: H Storage condition 25°C/60%RH

Testing Complete

W

(Contd)

Table 93. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: H Storage condition

25°C/60%RH

Testing Complete

Test End of life Storage Time Specification Initial 1 Month 3 Months 6 Months

(Contd)

Table 94. Stability data for DISKUS® 250/50 μg 60 Dose Product with overwrap

Packed Batch Number: H Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time Specification Initial 9 months 12 months 18 months

Table 95. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Packed Batch Number: H Storage condition

25°C/75%RH

Testing Complete

9s

(Contd)

Table 96. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Packed Batch Number: H Storage condition

25°C/75%RH

Testing Complete

Table 97. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH for 17 months with overwrap

Packed Batch Number: H Storage condition

25°C/7S%RH

Testing Complete

Test End of life Storage Time Specification Initial 2 Weeks 1 Month 1.5 Months

(Contd)

Table 98. Stability data for DISKUS® 250/50 μg 60 Dose Product without overwrap

Samples previously stored at 25°C/60%RH for 17 months with overwrap (cont'd)

Packed Batch Number: H Storage condition

25°C/75%RH

Testing Complete

Test End of life Storage Time

Table 99. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: I Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time

K)

©

(Contd)

Table 100. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: I Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time

Specification Initial I 9 Months I 12 Months 18 Months

(Contd)

Table 101. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: I Storage condition 25°C/60%RH

Testing Complete

Test End of life Storage Time Specification Initial 1 Month 3 Months 6 Months

κ> κ>

(Contd)

Table 102. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: I Storage condition 25°C/60%RH

Testing Complete

κ>

Table 103. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Packed Batch Number: I Storage condition

25°C/75%RH

Testing Complete

(Contd)

Table 104. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Packed Batch Number: I Storage condition

25°C/75%RH

Testing Complete

κ>

Ul

Table 105. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 17 months

Packed Batch Number: I Storage condition

25°C/75%RH

Testing Complete

κ>

(Coπt'd)

Table 106. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 17 months (cont'd)

Packed Batch Number: I Storage condition

25°C/75%RH

Testing Complete

-4

Table 107. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: J Storage condition 25°C/60%RH

Testing Complete

κ>

OO

(Cont'd)

Table 108. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: J Storage condition 25°C/60%RH

Testing Complete

κ>

(Contd

Table 109. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: J Storage condition 25°C/60%RH

Testing Complete

O

(Contd)

Table 110. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: J Storage condition 25°C/60%RH

Testing Complete

Table 111. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Packed Batch Number: Storage condition

25°C/75%RH

Testing Complete

(Cont'd)

Table 112. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Packed B atch Number: J Storage condition

25°C/75%RH

Testing Complete

Table 113. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 17 months

Packed Batch Number: J Storage condition

25°C/75%RH

Testing Complete

(Cont'd)

Table 114. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Samples previously stored at 25°C/60%RH with overwrap for 17 months (cont'd)

Packed Batch Number: J Storage condition

25°C/75%RH

Testing Complete

Table 115. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: K Storage condition 25°C/60%RH

Testing Complete

(Contd)

Table 116. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: K Storage condition 25°C/60%RH

Testing Complete

-4

(Contd)

Table 117. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: K Storage condition

25°C/60%RH

Testing Complete

Test End of life Storage Time

(Contd)

Table 118. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: K Storage condition 25°C/60%RH

Testing Complete

Table 119. Stability data for DISKUS® 500/5.0 μg 60 Dose Product without overwrap

Packed Batch Number: K Storage condition

25°C/75%RH

Testing Complete

(Contd)

Table 120. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Packed Batch Number: K Storage condition

25°C/75%RH

Testing Complete

Table 121. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 17 months

Packed Batch Number: K Storage condition

25°C/75%RH

Testing Complete

(Cont'd)

Table 122. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Samples previously stored at 25°C/60%RH with overwrap for 17 months (cont'd)

Packed Batch Number: K Storage condition

25°C/75%RH

Testing Complete

W

Table 123. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: L Storage condition 25°C/60%RH

Testing Complete

(Contd)

Table 124. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: L Storage condition 25°C/60%RH

Testing Complete

(Contd)

Table 125. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number' L Storage condition 25°C/60%RH

Testing Complete

(Contd)

Table 126. Stability data for DISKUS® 500/50 μg 60 Dose Product with overwrap

Packed Batch Number: L Storage condition 25°C/60%RH

Testing Complete

Table 127. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Packed Batch Number: L Storage condition

25°C/75%RH

Testing Complete

OO

(Contd)

Table 128. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Packed Batch Number: L Storage condition

25°C/75%RH

Testing Complete

VO

Table 129. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap Samples previously stored at 25°C/60%RH with overwrap for 17 months

Packed Batch Number: L Storage condition

25°C/75%RH

Testing Complete

O

(Coπt'd)

Table 130. Stability data for DISKUS® 500/50 μg 60 Dose Product without overwrap

Samples previously stored at 25°C/60%RH with overwrap for 17 months (cont'd)

Packed Batch Number: L Storage condition

25°C/75%RH

Testing Complete

CΛ

The invention has been described in reference to the embodiments set forth above. It should be appreciated that such embodiments are for illustrative purposes only, and do not limit the scope of the invention as defined by the claims.