WO2013091006A1 - Process for dry powder blending - Google Patents

Abstract

The present invention relates to processes for dry powder blending of microparticles, and to dry powder blends for use as pharmaceuticals. In particular, the present invention relates to a process for preparing a dry microparticulate powder mixture comprising a micronised drug and micronised pharmaceutically acceptable excipient, wherein the process comprises the step of fluidised blending of the micronised drug and the micronised pharmaceutically acceptable excipient. Said powder mixture may further be blended with a pharmaceutically acceptable carrier. The dry microparticulate powder mixture is intended for use in portable inhalation systems.

Description

PROCESS FOR DRY POWDER BLENDING

FIELD

[0001] The present disclosure relates, generally to processes for dry powder blending of microparticles and to dry powder blends for use as pharmaceuticals, particularly, but not exclusively, for use in portable inhalation systems.

BACKGROUND

[0002] The reference in this specification to any prior publication (or information derived from it), or to any matter which is known, is not, and should not be taken as an acknowledgment or admission or any form of suggestion that that prior publication (or information derived from it) or known matter forms part of the common general knowledge in the field of endeavour to which this specification relates.

[0003] Developing any efficient and effective portable inhalation system for medicinal use requires a complex integration of formulation and device technologies. In the case of each dry powder inhaler (DPI) system, each device technology is different, so creating a reliable and reproducible aerosol cloud composed of micron-sized particles provides its own unique set of significant technical and commercial challenges in the formulation development process.

[0004] DPI formulators face a basic powder technology challenge. The ultra-fine particle size which is required for effective deep lung delivery, typically less than 10 μπι, also inherently produces very cohesive powder characteristics. This property can make the micron-sized drug powders very difficult to handle in many stages of the DPI product life, including the manufacturing process, the dose metering process, through to the delivery from device to the patient, In addition, dosage amounts usually dictate that a diluent is required to ensure the dosage unit is suitable to be metered in a practical form. As a result, several formulation approaches are typically used to counter these challenges, the majority of which involve means of creating larger-sized powder units which consequently exhibit better powder flow properties.

[0005] The most common approach is to associate the micron-sized drug particles with larger (coarse) carrier particles (such as lactose), typically of a size in the region of 40 to 1000 um. The behaviour of these mixtures then becomes a complex function of the adhesive and cohesive forces present. These forces include the ubiquitous van der Waals forces as well as possible influences from electrostatic and capillary forces. However, he forces of adhesion and cohesion experienced by the drug particles should not be too strong as the drug particles should detach from the carrier particles during inhalation.

[0006] In order to address this, such formulations may also contain a third or ternary fine particle component. Typically this is milled lactose powders of size between about 1 and 20 μm which helps control the forces present and substantially improve the release of respirable drug from the larger carrier particles.

[0007] This property is generally measured by the parameter "Fine Particle Dose" (FPD) or "Fine Particle Fraction" (FPF) but it can be expressed in other ways. The FPD is the measured mass of the drug that has an aerodynamic diameter less than a specific size, and typically this is less than about 7 μτη. FPF is the measured mass percentage proportion of the total metered dose of the drug that has an aerodynamic diameter less than a specific size, and typically this is less than about 7 um. The addition of a microparticulate excipient, such as fine lactose, can lift FPD and FPF of drug in admixture with carriers, where FPF typically can be raised from the 5 to 10% level to in excess of 15 to 20%.

[0008] However, it is technically very difficult to produce the desired mix of a microparticulate drug and microparticulate excipient, such as fine lactose, to achieve such an advantageous random blend. This is because both components are inherently very cohesive in their original state. In order to mix well, the cohesive structures, i.e. agglomerates, must be destroyed. [0009] Current best practice used in all known commercially relevant processes for mixing a fine drug, fine excipient and coarse carrier, involves a blending process whereby the drug, fine excipient and coarse carrier are all blended together in a high shear blender, where materials are added together or at staggered times but in what is effectively a single blend step in a single unit process blender. Using the standard-type mechanical blenders, the process is conducted in a bulk batch process with a high shear applied to effect de- agglomeration and the required mixing of the cohesive fine components, in order to achieve suitable uniformity of content of the micronised drug within the blend. However, shear forces are applied violently and non-specifically in a localised region of the blender, usually at the base of the vessel, and inherently results in the uneven application of shear across the blend. As a result, excess energy must be applied to ensure that with the random movement, all parts of the mix are sufficiently blended. This generally means that some parts of the blend will receive repeated shear energy and hence substantially increases the level of particle damage, notably surface damage, disorder or even fracture especially to the coarser particles, resulting in unwanted changes in the particles and creating future physical and chemical instability issues. Furthermore, there is a compromise between the requirement of prolonged mixing and higher energy to ensure drug particles and fine excipient particles being sufficiently mixed, and the competing process of the fine drug particles instead becoming strongly attached to the larger carrier particle surface. According to one theory, a fine excipient facilitates the aerosolisation efficiency of the drug by forming loose agglomerates with the drug, hence minimising direct contact with the large carrier particle surface, but once the drug particle is attached and potentially pressed onto the large carrier particle surface by excess shear forces, this makes it more difficult to remove drug from the carrier in aerosolisation.

[0010] The above issues become even more difficult if more than one drug is being mixed, because every drug has a different balance of cohesive and adhesive forces with respect to the excipient components and consequently the rate of mixing and rate of transfer to the carrier particle surface will differ. [0011] There is a further critical inherent issue with heel residue (caking)/drug rich agglomerates-formed in this type of process, as a result of agglomerated drug particles getting stuck on mixer blades or in certain locations of the vessel before they can be sufficiently mixed. These agglomerates become fixed and are not further released into the mixing process, forming "hot spots" of drug. These "hot spots" result in a lack of blend uniformity and the consequential risk of batch failure, or batch inconsistency.

[0012] These processes also have the added logistical and cost burden of storage of the blended powder. Manufacturers may find it necessary to store traditionally blended powders at a specifically controlled temperature and humidity while the surface damage noted above can be subject to repair by controlled exposure to the moisture or dissipation of the electrostatic charge built up as a result of the high shear forces applied. These are processes known as "conditioning". This not only adds a time factor to production but also increases site storage requirements.

[0013] There is, therefore, a need for processes which allows for blending of a micronised drug, micronised excipient and a pharmaceutically acceptable carrier, and which may ameliorate or reduce one or more of the disadvantages discussed above.

SUMMARY

[0014] The present disclosure provides processes which allow for the preparation of powder compositions comprising a micronised drug, a micronised excipient and a carrier, in a two-step process. Thus, in some embodiments, there is provided a process for producing a dry powder mixture, comprising a micronised drug and a micronised pharmaceutically acceptable excipient_j comprising the step of fluidised blending of the micronised drug and excipient. The microparticles are blended together and de- aggleomerated by the application of a fluidised gas, blending advantageously being carried out in the substantial absence of larger particles such as carriers. The resulting microparticulate powder mixture may then be mixed or blended with a pharmaceutically acceptable carrier and, optionally, other suitable excipients. By eliminating the need for traditional high energy and high shear blending of micronised drug and excipient particles, typically carried out by blending with larger carrier particles in a single-step, 3 -phase blending process, embodiments of the present processes may substantially avoid or minimize size reduction and surface damage/morphology changes, not only to the micronised particles, but also to larger carrier particles, which might otherwise arise.

[0015] In one embodiment, there is provided a process for producing a dry powder mixture comprising a micronised drug and a micronised pharmaceutically acceptable excipient, said process comprising the step of fluidised blending of the micronised drug and micronised pharmaceutically acceptable excipient, in the substantial absence of a pharmaceutically acceptable carrier.

[0016] In one embodiment, there is provided a process for preparing a dry microparticulate powder mixture comprising a micronised drug and a micronised pharmaceutically acceptable excipient, wherein the process comprises the step of fluidised blending of the micronised drug and the micronised pharmaceutically acceptable excipient. In some embodiments, the microparticulate mixture may contain more than one drug and/or excipient, and may further contain one or more other pharmaceutically acceptable additives [0017] The resulting blended microparticulate dry powder mixture comprising the micronised drug and the micronised pharmaceutically acceptable excipient may then be further blended with one or more pharmaceutically acceptable carriers, and optionally one or more further phannaceutically acceptable excipients or additives.

[0018] Accordingly, in a further embodiment, there is provided a process for producing a dry powder composition comprising:

a) preparing a dry microparticulate powder mixture comprising fluidised blending of a micronised drug and a micronised pharmaceutically acceptable excipient; and

b) blending the powder mixture from step a) with a phannaceutically acceptable carrier.

[0019] In one or more embodiments, the micronised drug and the dry powder mixture and/or composition are suitable for administration to a patient by inhalation or oral routes. Thus, the powder composition may be administered as dry powder, e.g. by inhalation, or may be subjected to further processing, for example tableting or granulation for oral administration.

[0020] In some embodiments, the micronised pharmaceutically acceptable excipient is fine, or micronised, lactose.

[0021] In some embodiments, the pharmaceutically acceptable carrier is coarse lactose.

[0022] In further aspects, there is provided dry powder mixtures and dry powder compositions produced by the processes above. DESCRIPTION

[0023] Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise" and variations such as "comprises" and "comprising" will be understood to imply the inclusion of a stated integer or step or group of integers but not the exclusion of any other integer or step or group of integers or steps.

[0024] Throughout this specification and the claims which follow, unless the context requires otherwise, the phrase "consisting essentially of, and variations such as "consists essentially of will be understood to indicate that the recited element(s) is/are essential i.e. necessary elements. The phrase allows for the presence of other non-recited elements which do not materially affect the characteristics of the disclosure but excludes additional unspecified elements which would affect the basic and novel characteristics of the disclosure described herein.

[0025] The singular forms "a", "an" and "the" include plural aspects unless the context clearly dictates otherwise.

[0026] AH aspects, embodiments and examples described herein are encompassed by the term invention.

[0027] It has now been found that the application of a fluidised gas, in the absence of mechanical mixing parts such as screws, blades or paddles, to a mixture of the micronised drug and micronised excipient particles allows for gentle homogenisation and de- agglomeration of the micronised particles without the application of localized, prolonged and excessive energy. Advantageously, in some embodiments the fluidised blending process does not result in any significant particle size reduction and/or surface damage/morphology changes.

[0028] The processes may be carried out on a batch-wise or continuous basis. Advantageously, in some embodiments, by varying the flow rates and direction of the incoming gas jets, homogenous blending may be achieved for a particular micronised drug/excipient mix on a continuous basis.

[0029] Jet (or fluidised energy) mills are traditionally used for particle size reduction. Precisely aligned jets create a vortex into which a material is fed along an engineered tangent circle. High-speed rotation subjects the material to particle-on-particle impact,, creating smaller particles. While centrifugal force drives large particles toward the perimeter, the resulting fine particles move toward the centre where they exit. The present disclosure relates to the use of jet mills, used under conditions and suitably modified to provide acceleration and shear forces in the localised fluid-flow jets and lower energy impacts that are suitable to achieve de-agglomeration, and substantially avoid high energy impacts (which result in substantial particle size reduction and particle surface damage), in fluidised blending of the micronised drug and micronised excipient components.

[0030] As used herein, "fluidised blending" refers to the blending of powder components (e.g. microparticles) by the application, of one or more jets or streams of a gas or gaseous fluid such that the powder particles are blended and de-agglomerated but wherein no substantial particle size reduction by attrition or milling and/or surface damage occurs.

[0031] In some embodiments, fluidised blending can be performed in a jet mill apparatus, by the application of an inwardly directed circular or spiral air flow of a gas or gaseous fluid from one or multiple points emanating inwards at angles from tangential or greater angles from the internal circumferential walls of a substantially circular vessel or chamber and at a pressure which avoids or minimizes high impact collisions between the particles and which result in particle size reduction and damage. In further embodiments, the blending occurs in the absence of mechanical blending or grinding parts such as arms, paddles and blades, and in the substantial absence of particle size reductions or consequential decrease in FPF or FPD.

[0032] The gas may be any suitable gas for use with the particular drug and excipient. In some embodiments the gas is air, but in other embodiments the gas may be nitrogen, oxygen, helium, or argon, either as an essentially pure gas or a mixture of gases, and may contain a mix of vapours including a range of water or solvent vapours. In some embodiments the gas is a dry gas i.e. substantially free of water vapours. The injecting (feed) and "grinding" gases may be the same or different, but in some embodiments are the same.

[0033] The fluidised blending is advantageously carried out in a substantially circular vessel or chamber having a central exit port, and jet nozzles, which blow gas or gaseous fluid into the chamber, are distributed around the circumference, thereby generating and maintaining a vortex. The jet nozzles are positioned such that jets of gas or gaseous fluid may be blown into the chamber substantially tangentially e.g. a few degrees to the tangent, or at angles greater than a tangent. It will be understood that the angles of the jets may equally be expressed as angles to the perpendicular. For example, an angle of 60° to the tangent, is can also be expressed as an angle of 30° to the perpendicular. The angles of the jets are such that high energy particle-particle impacts are minimised so as to substantially avoid particle size reduction and/or surface damage. In certain embodiments, these angles are less than or about 60° to the tangent, such as less than or about 45°. In some further embodiments, the angle may be, less than or about 30°, less than or about 25°, less than or about 20°, 15° or 10° to the tangent. Centripetal forces direct the smaller (e g. de- agglomerated) particles towards the central exit port through which they are removed from the vessel or chamber while centrifugal forces direct the larger agglomerates to the perimeter. In this manner, small, i.e. de-agglomerated particles may easily and rapidly exit the vessel, after one or a small number of spiral trajectories, and after, for example, a few hundredths of a second residence, but with the large particles remaining in the vessel until they become sufficiently de-agglomerated to be small enough to exit. In some embodiments, the process may therefore advantageously lend itself to a continuous feed/exit process.

[0034] The micronised particles (drug and excipient) to be blended can be fed into the chamber together, via a one or more feed line(s), or individually, via separate lines, at a rate which allows for the desired blend. Thus, in one embodiment, the micronised drug and cxcipient are separately fed into the chamber. In other embodiments, the micronised particles (microparticles) are first subjected to a pre-mix step, for example in a tumbler, and the pre-mixx is then fed into the chamber for fluidised blending.

[0035] In each case, the feed rate can be optimised to be within a range, which can be identified by trial evaluation, such that the particle concentration within the vessel is high enough to provide sufficient numbers of collisions, but is not so high that the particle concentration over-saturates the chamber, such that particles have inadequate space between each other to allow acceleration, development of shear forces and effective collisions to occur. A minimum free path in the flow is desired to effect such forces and collisions, as can be determined by the skilled person.

[0036] The pressure of the jet gas. (also referred to herein as the "grinding" gas, although it is to be understood that no substantial grinding or particle attrition occurs) is such that high energy impacts between the particles are avoided, and thus are lower than those typically used when particle attrition is intended, i.e, milling. Thus, in some embodiments, the pressure of the jet gas is about 2.5 bar or less. In further embodiments, the pressure of the jet gas is, about 2.0 bar or less, or 1.5 bar or less, or 1.0 bar or less or about 0.5 bar. In further embodiments, the jet gas pressure is about 1.0 bar.

[0037] The feed (or injecting) gas delivers the microparticles to the mixing chamber oi vessel. The pressure of the feed gas, may the about the same as the grinding gas or may be higher or. lower. In some embodiments, the pressure of the feed gas is about 2.5 bar or less, for example about 2.0 bar or about 1.5 bar or less; or about 1.0 bar, or less or about 0.5 bar. In further embodiments, the pressure of both the feed and jet gases is about 1.0 bar.

[0038] The feed rate may be adjusted to suit the amount of microparticulate powder to be blended and will depend on the diameter of the fluidisation chamber. For example, for a 50 mm chamber, a suitable feed rate may be in the range of about 5-20 g/min. Exemplary feed rates in this range include about 5-7 g/min, about 8-9 g/min, about 10-12 g/min, about 13-15 g/min, and about 16-18 g/min. Higher feed rates for example in the range of 20-100 g/min may be used for chambers of 100mm or more. Suitable feeders include screw feeders and vibratory feeders. In some embodiments, the feeder is a screw feeder.

[0039] As used herein, the term "micronised", or "fine", "microparticle" or "microparticulate", when used with reference to a drug, refers to drug particles having a size, measured as mass median diameter, of about 20 um or less, such as 10 μπι or less, or about 7, 6, 5, 4 or 3 am or less.

[0040] As used herein, the term "micronised", or "fine", "microparticle" or "microparticulate" when used with reference to a pharmaceutically acceptable excipient, refers to excipient particles having a size, measured as mass median diameter, of about 20 μπι or less, such as 10 μπι or less, or about 7, 6, 5, 4 or 3 μπι or less.

[0041] A "microparticulate powder mixture", as used herein, refers to a powder mixture consisting essentially of microparticles, i.e. particles having a mass median diameter of about 20 μπι or less, such as 10 um or less, or about 7, 6, 5, 4 or 3 μηι or less. Thus, in the blending of such a mixture, larger particle components, such as carrier particles, are substantially absent.

[0042] In some embodiments, the micronised drug particles have a mass median diameter less than that of the excipient particles. In further embodiments, the drug particles have a mass median diameter of about 5 to 3 μτη or less and the micronised excipient particles have a mass median diameter in the range of about 10 to 7 μπι.

[0043] The mass median diameter (being the diameter at which 50% by mass of the particles have a diameter greater and 50% by mass have a diameter smaller) can be measured by any suitable means, such as laser light scattering (e.g. using a Malvem

Mastersizer 2000 instrument).

[0044] The FPD can be , specified as the mass of the metered dose of the drug that has a mass median aerodynamic diameter less than about 7 um, such as less than about 5 or 3 urn. FPF can be specified as the mass percentage proportion of the metered dose of the drug that has an aerodynamic diameter less than about 7 μηι, such as less than about 5 or 3μm..

[0045] The mass median aerodynamic diameter can be measured by an impinger or pharmacopeia impactor method as defined by the US Pharmacopeia, by any suitable means such as an Andersen Cascade Impactor (ACI), Marple Miller Impactor or Next Generation Impactor (NGI).

[0046] Reference to "no significant", "minimal" or "substantially no" particle size reduction or change, and the like, is intended to convey that that the blended powder mixture exhibits no significant, minimal or substantially no change in a measureable quantity reflecting particle size, such as, for example, mass median diameter as determined by laser diffraction or scathing.

[0047] The micronised drug and excipient may be brought into a micronised state (individually or together) by any suitable means, including, but not limited to, spray drying and attrition means, such as air jet milling, spiral air jet milling, fluid-bed jet milling, ball milling, pearl milling, bead milling or other alternative forms of milling, such as pin milling, end-runner milling or centrifugal milling. Any form of controlled precipitation can be used, including crystallisation, such as crash crystallisation, solvent-antisolvent methods, all known supercritical fluid based particle engineering techniques, ultrasound- based crystallisation techniques, emulsion based particle engineering techniques, spinning disc based particle synthesis techniques, microfluidic based precipitation techniques, and any other particle engineering techniques. In certain embodiments, the microparticles are prepared by a dry, non-solvent-based, milling process.

[0048] Pharmaceutically acceptable carriers and micronised excipients contemplated herein include, mono, di- and polysaccharides, for example, sugars such as glucose, lactose fructose, sucrose, mannitol, xylitol, sorbitol, dextran, and trehalose, starches, such as corn starch, and cellulose and derivatives thereof, such as ethyl cellulose. It will be appreciated that the carrier and excipient are independently chosen, that is to say, the carrier and excipient may be the same compound or different.

[0049] In some embodiments, the pharmaceutically acceptable excipient is micronised lactose.

[0050] The size (diameter) of the pharmaceutically acceptable carrier particles is substantially greater than that of the micronized drug and pharmaceutically acceptable excipient, typically having a mass median diameter of at least about 30, 40, 50, 60, 70, 80, 90 or 100 μm In some embodiments, the pharmaceutically acceptable carrier is lactose. In further embodiments, the carrier is lactose, having a mass median diameter of at least about 30-1000 μm (coarse lactose). In further embodiments, the carrier has a mass median diameter in a range selected from about 40 to about 200 μιη, or about 40 to about300 μηι, about 40 to about 400 μηι, about 40 to about 500 um or about 40 to about 600 μηι. The mass median diameter can be measured by laser light scattering.

[0051] The coarse lactose may be conventional lactose or an agglomerated lactose cluster. Conventional lactose consists of primary particles typically in the size range of about 40 or 50 to 100 μm . Due to its crystal structure and slip planes, conventional lactose takes a "tomahawk" shape. A less conventional approach to inhaled product formulation is to use agglomerated lactose cluster particles which comprises an agglomerate of single lactose Crystals, arranged in an irregular and jagged shape with a typical particle size greater than 150μm. Without limiting the disclosure by theory, it is suggested that when blended with micronised drug and fine excipient particles the irregular shape of the agglomerated lactose cluster particles may retain a higher level of the fine particles on its surface than conventional lactose. The reduction of loose fine and cohesive material from the use of agglomerated lactose clusters may serve to both create a more uniform blend and improve product performance by enabling the release of a higher percentage of fine drug particles beyond the oropharyngeal region. In addition, "cluster blends" may also exhibit more free- flowing properties to conventional lactose blends due to this reduction. Thus, in some embodiments, the coarse lactose is an agglomerated cluster lactose having particles with a mass median diameter in the range of about ISO to 1000 μηι, such as about 150 to 700 μηι.

[0052] In some embodiments both the micronised excipient and the pharmaceutically acceptable carrier are lactose.

[0053] The micronised drug includes any physiologically active agent intended for the treatment, prevention, cure or diagnosis of a disease or condition, or symptom thereof, and which is desired to be blended with micronised excipients and carriers for administration.

[0054] In some embodiments, the drug is one which is intended for pulmonary administration via inhalation. Some non-limiting examples thereof include B2 adrenergic receptor agonists {e.g., salbutamol (e.g. sulphate), salmeterol, formoterol, salmefamol, fenoterol, terbutaline, albuterol and their pharmaceutically acceptable salts), antibiotics (e.g. erythromycin, azithromycin, tobramycin, clarithromycin, gentamicm, ceftazidime, piperacillin, ciproflozacin and rifampicin), antivirals, antifungals, mucolytics, bronchodilators, anti-inflammatories (e.g., steroidal anti-inflammatories, such as methyl prednisolone, prednisolone, dexamethasone, fluticasome propionate, budesonide, flunisolide, rofleponide, ciclesonide, butixocort and pharmaceutical esters and salts; NSAIDS), glucocorticoids, anti-muscarinc agents (e.g., ipatropium, tiotropium and glucopyrrolate and their salts), and anti-cholinergics (e.g., carbinoxamine maleate, clemastine fumarate, diphenylhydramine hydrochloride, acrivastine, astemizole, levocabastine hydrochloride, terfenidine, fexofenadine) and the pharmaceutically acceptable derivatives, esters or salts thereof

[0055] In further embodiments, the micronised drug particles consist of active agent only, for example are non-coated particles.

[0056] It will be appreciated that although in some embodiments described herein the dry powder composition is intended for administration by inhalation, the processes disclosed herein are applicable to the blending of microparticulate dry powders intended for administration by means other than inhalation (e.g. Oral administration). The blended mixtures and compositions may then be subjected to further processes such as tableting and granulation to provide oral dosage forms. Suitable processes are known in the art and are described in numerous reference texts, for example Remington's Pharmaceutical Sciences, 2 V Edition.

[0057] Furthermore, the fluidised blending of micronised components to form a micronised dry powder mixture may include one or more drugs and/or one or more excipients. It is also envisaged that one or more dry powder mixtures prepared by fluidised blending may be further blended with one or more pharmaceutically acceptable carriers. Thus, the disclosure herein also contemplates the further inclusion of one or more additional drug and/or excipient components, and/or additional carrier components. In some embodiments, additional micronised drug and/or micronised excipient components will be blended in a first step as described herein and one or more additional coarse carrier components can be blended in a further step as described herein.

[0058] In some embodiments, the processes, mixtures and compositions described herein may optionally comprise further pharmaceutically acceptable additives, including binders, disintegrants, lubricants, glidants, preservatives, flavourings, colourings, sweeteners, diluents and bulking agents, which may be added, as appropriate, in micronised form, during the fluidised blending step, with the pharmaceutically acceptable carrier during the second step.

[0059] The processes and powders of the disclosure will now be further described with reference to the following examples which are included for the purpose of illustrating some embodiments but are not intended to limit the generality hereinbefore described. EXAMPLES

Example 1

[0060] The blending of a number of dry powder compositions was examined.

[0061] Compositions comprising:

1% (w/w) salbutamol sulphate

19% (w/w) fine lactose (Sorbolac 400);

80% (w/w) coarse lactose (Prismalc 40, Meggie GmbH or Respitose SV003, DMV- Fonterra);

were prepared using a number of blending strategies as outlined below. Table 1-1 sets out the physical properties of the powders used.

Table 1-1: Material properties of powders used in Example 1.

*(Do.i, (or D10) is the spherical equivalent diameter where 10% by mass of the particles of the powder has a smaller diameter, and by inference 90% has larger diameter.

Do s: is the spherical equivalent diameter where 50% by mass of the particles of the powder has a smaller diameter, similarly 50% is larger. The D0.5, (or D50) is also known as the mass median (spherical equivalent) diameter.

D0.9, is the spherical equivalent diameter where 90% by mass of the particles of the powder has a smaller diameter, and by inference 10% is larger.) [0062] Fluidised blending was conducted in a spiral jet mill (50ASMUI, Hosokawa Micron Corp., Japan) modified such that the four jets are directed at 66° from the perpendicular (i.e. 24° from the tangent). More traditional high shear blending was conducted in a mechanofusion ultra-high shear mixer (Nobilta Hosokawa Micron Corp., Japan) or a traditional high shear mixer (KG5-1L, Key International, USA). Tumbled lower shear blends were made in a Turbula T2 ( Glen Mills Inc., USA). The following abbreviations are used: J (Jet mill), Mechanofusion (M), High Shear Mixer (K) and Tumbler (T).

[0063] The blends were made as follows:

1: J2R (Jet, 2-step, Respitose)

[0064] The first mix of total mass 50g comprising micronised salbutamol sulphate and fine lactose (Sorbolac 400) mixed in a ratio 1:20, by mass was prepared by tumbling for 5 min prior to blending to ensure consistency in results. This blend was then fed into the 50AS jet mill at a feed rate of approximately 1 Og/min using a screw feeder, with air flow set as 2 bar (g) to the feed and 1 bar (g) to the fluidising jets.

[0065] This powder was recovered from the bag collector of the system. The second mixing step then comprised 50 g of powder mixed in ratio 20:80 of the fine preblend from the first mix and the carrier Respitose SV003. This was mixed in the Turbula T2 for 5 minutes at 72 rpm, and then recovered.

2: J2P (Jet, 2-step, Prismalac)

[0066] A procedure identical to that used to create (1) was used to create J2P, but the preblend from (1), was alternatively mixed in the Turbula T2 with the carrier Prismalac 40 instead of the carrier Respitose SV003.

3: M2R (Mechanofusion, 2-step, Respitose)

[0067] The first mix of total mass 30g comprising micronised salbutamol sulphate arid fine lactose (Sorbolac 400) mixed in a ratio 1:20, by mass was prepared by tumbling for 5 min prior to blending to ensure consistency in results. This blend was then mixed in the Hosokawa AMS Nobilta system at 600rpm for 5 minutes.

[0068] This powder was recovered from the Nobilta vessel. The second mixing step then comprised 50 g of powder mixed in ratio 20:80 of the fine preblend from the first mix and the carrier Respitose SV003. This was mixed in the Turbula T2 for 5 minutes at 72 rpm, and then recovered.

4: M2P (Mcchanofusion, 2-step, Prismalac)

[0069] A procedure identical to that used to create (1) was used to create M2P, but the preblend from (1), was alternatively mixed in the Turbula T2 with the carrier Prismalac 40 instead of the carrier Respitose SV003.

5: MIR (Mcchanofusion, 1-step, Respitose)

[0070] 30g comprising micronised salbutamol sulphate, fine lactose (Sorbolac 400) and Respitose SV003 in a ratio 1:19:80, was mixed in the Hosokawa AMS Nobilta system at 600rpm for 5 minutes.

6: M1P (Mechanofusion, 1-step, Prismalac)

[0071] 30g comprising micronised salbutamol sulphate, fine lactose (Sorbolac 400) and the carrier Prismalac 40 in a ratio 1 ; 19:80, was mixed in the Hosokawa AMS Nobilta system at 600rpm for 5 minutes.

7: T2R (Tumber, 2-step, Respitose)

[0072] The first mix of total mass 50g comprising micronised salbutamol sulphate and fine lactose (Sorbolac 400) mixed in a ratio 1 :19, by mass was prepared by tumbling for 25 min in a Turbula T2 at 72 rpm.

[0073] This powder was recovered from the mixing vessel. The second mixing step then comprised 50 g of powder mixed in ratio 20:80 of the fine pre-blend from the first mix and the carrier Respitose SV003. This was mixed in the Turbula T2 for 5 minutes at 72 rpm, and then recovered.

8 : T2P (Tumber, 2-step, Prismalac)

[0074] The first mix of total mass 50g comprising micronised salbutamol sulphate and fine lactose (Sorbolac 400) mixed in a ratio 1 : 19, by mass was prepared by tumbling for 25 min in a Turbula T2 at 72 rpm.

[0075] This powder was recovered from the mixing vessel. The second mixing step then comprised 50 g of powder mixed in ratio 20:80 of the fine pre-blend from the first mix and the carrier Prismalac 40. This was mixed in the Turbula T2 for 5 minutes at 72 rpm, and then recovered.

9: TIR (Tumber, 1-step, Respitose)

[0076] 50g comprising micronised salbutamol sulphate, fine lactose (Sorbolac 400) and the carrier Respitose SV003 in a ratio 1:19:80, was mixed in the Turbula T2 at 72rpm for 45 minutes.

10: TIP (Tumber, 1-step, Prismalac)

[0077] 50g comprising micronised salbutamol sulphate, fine lactose (Sorbolac 400) and the carrier Prismalac 40 in a ratio 1:19:80, was mixed in the Turbula T2 at 72rpm for 5 minutes.

11: KIR (High Shear, 1-step, Respitose)

[0078] SOOg comprising micronised salbutamol sulphate, fine lactose (Sorbolac 400) and the carrier Respitose SV003 in a ratio 1 :19:80, was mixed in the KG5-1L high shear blender at 255rpm for 6 minutes. Comparison of Mixers Used

[0079] The Turbula T2F tumbler mixer has several different arm rotation speeds and can vary in mixing intensity. Comparatively, tumblers are considered a low intensity mixer that does not risk damage to the blended particles. This mixer was chosen primarily as a secondary mixer but also served as a base line comparative tool for the performance of the other mixers.

[0080] The high shear mixers served as a replicate for current batch blending technologies and serves as a comparison for the performance of blends from the other mixers.

[0081] Whilst different to traditional mechanofusion processes, the Nobilta mechanofusion process (Hosokawa Micron Corp., Japan) is not dissimilar in immediate appearance to a standard high shear impeller mixer. However, the major difference is that the impellers are designed to rotate relatively faster and compress and shear powder at the interface between the impeller and mixer walls in a similar manner to other mechanofusion systems. A very small clearance between the rotor blades and vessel walls (~lmm) allows high compaction of powder against the vessel walls, mimicking a traditional mechanofusion process. During operation powder is not only forced to the walls of the mixer where it is compressed and sheared but powder is lifted from system walls through the oscillating pathway of the bulk powder and under impeller blades.

Blend analysis

[0082] The blended compositions were left under laboratory ambient conditions (between 20-25°C and about 30-60% relative humidity) for at least 10 days. Blends were tested for content uniformity of salbutamol and fine particle dose using Cascade Impaction.

Blend content uniformity of salbutamol by HPLC

[0083] 50 mg of formulation was sampled from each blend pot by pouring contents of the sample pot in a line out on a foil sheet and taking powder from seven cross-sections along the powder line. [0084] Sampled quantities were placed in pre-weighed containers and then re-weighed to obtain released powder mass, weighing to 4 decimal places (New Classic MS, MS205DU, Mettler Toledo). Sample vials were placed in a tray and covered with aluminium foil until prepared for HPLC analysis. Once ready to use, vials were removed from tray and the dry powder rinsed into a 50mL volumetric flask, made to volume with ultra-pure water and filtered (Nylon filter, 0.45μπι, Restek) in preparation for HPLC analysis.

HPLC analysis

[0085] Salbutamol sulphate standard solutions (1.6 ± 0.16% w/v) were made by dissolving salbutamol sulphate reference standard (99.4% purity) in ultra-pure water.

[0086] HPLC was performed on a Prominence 2QA-Series, Shimadzu Corp, Japan instrument using an HPLC column (Symmetry C8, 5 urn, 15cm x 3.9mm ID) operated for 7 min on main run (flow rate 1.0 mL/min) at ambient temperature with a sample injection of 100μL, Detector (Prominence SPD-M20A, Shimadzu Corp, Japan) operated at 220nm (UV). HPLC Mobile phase comprised Buffer solution pH 3.7 (0.287% w/v sodium heptanesulphonate, 0.25% w/v potassium dihydrogen orthophosphate adjusted with 10% v/v orthophosphoric acid) mixed with acetonitrile in a volume ratio of 78:22.

Fine particle dose by cascade impactor

[0087] Prior to analysis, sets of 30 capsules were each filled with approximately 26mg ±1 mg of each blend using a automated capsule filling device (Quantos QUI, Mettler Toledo, USA) in order to standardise fill method.

[0088] Mass median aerodyndamic diameter was measured using an Andersen Mk II Cascade Impactor (ACI, Andersen Instruments Inc., Electron Corp.), prepared by washing the full device with HPLC mobile phase followed by acetone to ensure it was clean, and allowing to dry for 20 minutes. Collection plates were sonicated in hexane for 5 minutes, washed in acetone and allowed to dry. Once dry, each plate was immersed in plate coating solution (1% w/v silicone oil in hexane) and dried again. Rotahaler (provided by Glaxo SmithKlim) devices were washed with HPLC mobile phase and allowed to dry prior to use. The Impactor was assembled once all items were dry and inspected for blockages. Temperature and relative humidity were recorded before each preparation.

[0089] An empty Rotahaler device was inserted into the rubber mouthpiece and a lOOmL volumetric flask placed at the base to ensure complete recovery. Rotacaps were placed carefully into the Rotahaler device (body first with Rotacap flush with entrance of hole). The method ensured that powder, was present predominantly in the section inserted first. The rear portion was twisted through lSO'to break the Rotacap. After capsule breakage.the pump was operated for 5 seconds at an inlet air flow rate of 60L/min.

[0090] Once 10 Rotacaps had been successfully aspirated through device the cascade impactor was carefully dismantled and each stage and piece rinsed carefully with HPLC mobile phase into 50 mL (Throat + mouthpiece, Rotahaler, Rotacaps , Stages 1-7, filter) and 250mL (Pre-separator + Stage 0) volumetric flasks.

[0091] HPLC analysis was conducted as per content uniformity testing (Blend content uniformity of salbutamol by HPLC) and the procedure was duplicated for each blend.

Stability tests by cascade impactor

[0092] Approximately 5.5 months (165 days) after the formulations were^' created, fresh powder was extracted from the foil -sealed blend drums and used to fill 30 capsules as per previous cascade impactor tests. Testing methods were as per previous cascade impactor tests.

Content uniformity

[0093] Content uniformity was determined by HPLC. Through analysis of seven samples from each blend the mean salbutamol content and relative standard deviation (%RSD) for each mixture were determined (Table 1-2). Primary assay shows mean salbutamol content ^g) per 26mg of sample and the relative standard deviation (%RSD) across samples. Secondary assay shows the mean percentage of salbutamol per capsule compared to the theoretical label claim of 2^g/capsule.

Table 1-2: Content uniformity results for each blend.

Effect of blender on label claim and uniformity

[0094] For both Respitose and PrismaLac blends, a similar hierarchy is observed of both uniformity (decreasing %RSD) and mean salbutamol content.

Content uniformity

[0095] Jet mill (J) >Mechanofusion (M) > High Shear Mixer (K) > Tumbler (T). Mean salbutamol content (yield):

[0096] Jet mill (J) > High Shear Mixer (K) >Mechanofusion (M) > Tumbler (T). [0097] It is worth noting that whilst the jet mill and high shear mixer afforded a similar yield (87.0 % and 86.4% for Respitose blends respectively), the uniformity of the jet mill blend is superior (10.0% compared to 19.4%).

[0098] For both single and 2-step tumbler (T) blends, yield and uniformity were found to be well outside the acceptable limits (RSD < 5.0%) and highly variable (RSDs between 20.0% and 122.1%). Tumbled only blends containing PrismaLac were found to have drug- rich pockets with one or two samples from both the TIP and T2P blends containing between 4.9 and 7.6 times the specified drug content. It is believed that drug-rich pockets also exist in the Respitose blends as yields are extremely low (62.6% and 63.8%). micronised

Fine particle dose (FPD)

[0099] Through testing of sets of 10 capsules using the Andersen Cascade Impactor, distributions of salbutamol for several blends were determined.

Table 1-3: Cascade Impaction data for jet mill (J), mechanofusion (M) and high shear mixer (K) formulations. Columns contain mean salbutamol recovery, delivered dose, fine particle dose (FPD) per capsule (26 mg) and fine particle fraction (FPF) (fraction of salbutamol per capsule recovered in fine particle region).

Observations

[0100] It was observed immediately after production that the working area of the jet mill contained a small amount of deposition around the rim and outlet, with a slight build-up on the ceiling and feed inlet. Much of the deposition observed was found to be very loose and easily removed by tilting the grinding plate. As almost all of the powder was easily removed this form of deposition can be classed as residual or mixer holdup and is expected to be shifted by fresh powder during operation, essentially creating a self-cleaning device. This observation of only light, non-compacted deposition suggests that fluidised blending may avoid or minimize batch inconsistency due to build up or caking and could be utilized in a continuous, rather than a batch process.

[0101] In contrast, the high shear mixer was found to have irregular caking on the walls and lid of the vessel. Further caking (heel residue) was observed when the bulk of material was removed. It was observed that caking on mixer walls and base were quite well compacted and stable, and did not fall off or disintegrate during removal of the bulk material. This suggests that powder in some regions of the vessel underwent either poor or no mixing. Such regions are also commonly described as dead zones and are a common occurrence in traditional high shear mixers. The impeller blades were also found to have a small level of stable caking on the upper lip.

[0102] After blending of the primary blend, the mechanofusion mixer was found to have severe caking on the outer walls and mixer lid, and a stable thin film of powder on the impeller shaft and blades. When the bulk of material was removed much of the caking was found to require scraping to remove from the mixer and impeller surfaces. The wall caking was observed to be as thick as the clearance between the impeller blades and mixer walls (lmm) and is believed to be a compressed mass of poorly mixed powder. The caking of the mechanofusion primary blend is believed to be quite firm due to the physical properties of the fine lactose. As the blend consists of only highly cohesive materials of small particle size (salbutamol sulphate and SorboLac) the strong compression forces on the powder between impeller and walls were able to create a cake of high density and compaction with strong inter-particle and particle-wall adherence.

Stability Tests

[0103] Through testing of sets of 10 capsules using the Andersen Cascade Impactor, distributions of salbutamol for stored J2R, J2P and KIR blends were determined (Table 1- 4).

Table l-4:Cascade Impaction stability data for jet mill (J) and high shear mixer (K) formulations. Columns contain mean salbutamol recovery, delivered dose and fine particle dose (FPD) per capsule (26 mg) and fine particle fraction (FPF) (fraction of salbutamol per capsule recovered in fine particle region).

[0104] The jet mill blends showed superior results compared to the high shear blend. Example 2

[0105] lwt% salbutamol sulphate/commercially available milled lactose Lactohale LH230 (bulk density = 0.32 g/cm3, d oVd 0.50/d 0.90 = 3.1 / 10.4 / 24.3 microns respectively) blends were made, using the same modified jet mill as for Example 1 , at three different feed rates and jet (grinding) pressures with feed pressure held constant

[0106] A screw feeding system was used to feed the raw materials into the mixing chamber. [0107] Blending was performed at five different "grinding" pressures, a single feed pressure and three different feed rates. The results are set out in Table 2-1.

Table 2-1: Experimental matrix for salbutamo) sulphate screw feeder blending.

[0108] Jet pressures were capped at 2.5 bar (g) as this was the highest pressure at which minimal milling of the fine lactose occurred and above which increasing size reduction occurred (as determined by dry powder laser diffraction).

[0109] The fine product was collected and tumbled with coarse lactose (Respitose SV003) in a ratio of 1:4. Methods of uniformity and inhalation property testing were the same as previous examples, except that 16 measurements were taken for uniformity and not 10.

[0110] All inhalation measurements for this example were conducted with a standard commmercially available GSK Ventolin Rotahaler.

Content uniformity

[0111] The accepted criteria for the uniformity of the commercial formulation is an RSD of 5%. It was found that all blend manufactured fell substantially below this level, with many below 2% (Table 2-2). This finding demonstrates the effectiveness of this method for producing a high degree of content uniformity in pharmaceutical powder blends, even at low drug concentration.

Table 2-2: %RSDs for each blend manufactured.

[0112] In addition to the fifteen blends manufacture, three additional blends were manufactured which show that satisfactory uniformity can still be achieved when less well controlled feeding systems (e.g. vibratory feeder) are employed, an excessive feed rate is used or there is no pre-tumbling (instead, using two separate screw feeding systems) of drug and fine lactose prior to feeding (Table 2-3).

Table 2-3; Content uniformity results for other blends

Inhalation properties (DD/FPD/FPF)

[0113] In addition to content uniformity measurements, delivered dose (DO), fine particle mass (FPD) and fine particle fraction (FPF) were measured for each blend using the same method as the earlier examples and compared to that obtained for a control blend prepared in the traditional single-step manner.

[0114] It was observed that all blends gave higher fine particle doses and fractions than the control blend, with many offering FPFs up to 10% higher than the control.

Table 2-4: Initial inhalation data for blends

[0115] It was also found that other blends tested in this study showed superior inhalation properties compared to the control blend. While the drug content of the blend and hence DD was slightly lower, the FPD was consistently higher.

3-month stability.

[0116] The inhalation properties of the blends were tested after 3 months of storage in a controlled relative humidity (65%) and temperature (35°C). It was found that all blends decreased slightly in FPF, but decreases were less than is previously experienced using conventional blending, and all blends had higher (or equal in the case of CI) values than the initial FPF for the commercial formulation. The results are depicted in Table 2-5.

Table 2-5: 3-moath stability data for FCD blends

Claims

1. A process for preparing a dry microparticulate powder mixture comprising a micronised drug and a micronised pharmaceutically acceptable excipient, wherein the process comprises the step of fluidised blending of the micronised drug and the micronised pharmaceutically acceptable excipient.

2. A process for producing a dry powder composition comprising:

a) preparing a dry microparticulate powder mixture comprising fluidised blending of a micronised drug and a micronised pharmaceutically acceptable excipient; and

b) blending the powder mixture from step a) with a pharmaceutically acceptable carrier,

3. The process according to claim 1 or 2 wherein the micronised drug has a mass median particle size of about 20 μm or less.

4. The process according to claim 3 wherein the drug has a mass median particle size of about 7μm or less.

5. The process according to any one of claims 1 to 4 wherein the drug is selected from the group consisting of B₂ adrenergic receptor agonists, salbutamol, salbutamol sulphate, salmeterol, formoterol, salmefamol, fenoterol, terbutaline, albuterol and their pharmaceutically acceptable salts, antibiotics, erythromycin, azithromycin, tobramycin, clarithromycin, gentamicin, ceftazidime, piperacillin, ciprofloxacin, rifampicin, antivirals, antifungals, mucolytics, bronchodilators, anti-inflammatories, steroidal antiinflammatories, methyl prednisolone, prednisolone, dexamethasone, fluticasome propionate, budesonide, flunisolide, rofleponide, ciclesonide, butixocort, non-steroidal anti-inflammatories, glucocorticoids, anti-muscarinic agents, ipatropium, tiotropium, glucopyrrolate, anticholinergics, carbinoxamine maleate, clemastinefumarate, diphenylhydramine hydrochloride, acrivastine, astemizole, levocabastine hydrochloride, terfenidine, fexofenadine, and the pharmaceutically acceptable derivatives, esters or salts thereof.

6. The process according to any one of claims 1 to 5 wherein the micronised excipient has a mass median particle size of about 20 μιη or less.

7. The process according to any one of claims 1 to 6 wherein the excipient is selected from the group consisting of glucose, lactose fructose, sucrose, mannitol, xylitol, sorbitol, dextran, trehalose, starches, and cellulose and derivatives thereof.

8. The process according to claim 7 wherein the excipient is lactose.

9. The process according to any one of claims 1 to 8 wherein the fluidised blending is conducted with a jet gas pressure of 2.5 bar or less.

10. The process according to claim 9 wherein the jet gas pressure is about 1 bar.

11. The process according to any one of claims 1 -10 wherein the fluidised blending is conducted with a feed pressure of 2.5 bar or less.

12. The process according to claim 11 wherein the feed pressure is about 1 bar.

13. The process according to any one of claims 2 to 12 wherein the wherein the carrier has a mass median particle size of about 30 urn to about ΙΟΟΟμπι.

14. The process according to any one of claims 2 to 13 wherein the carrier is selected from the group consisting of glucose, lactose fructose, sucrose, mannitol, xylitol, sorbitol, dextran, trehalose, starches, and cellulose and derivatives thereof.

15. The process according to claim 10 wherein the carrier is lactose.

16. The process according to claim 11 wherein the lactose is conventional lactose and has a mass median particle size in the range of about 50 to 100 um.

17. The process according to claim 10 wherein the lactose is clustered lactose having agglomerates with a mass median particle size greater than about 150 μηχ