1
`
`3M COMPANY 2010
`Mylan Pharmaceuticals Inc. v. 3M Company
`IPR2015-02002
`
`

`
`Library of Congress Cataloging-in—Puhlieation Data
`
`Inhalation aerosols : physical and biological basis for therapy I edited by
`Anthony J. Hickey.
`p. cm. — (Lung biology in health and disease ; v. 94)
`Includes bibliographical references and index.
`ISBN 0-324 7-9702-7 (hardcover : alk. paper)
`[ DNLM:
`II. Series
`I. Aerosol therapy.
`I. Hickey, Anthony I.
`I. Aerosols — therapeutic use. 2. Aerosols — pharmacology.
`3. Respiratory Airflow — physiology. W1 LU62 v. 94
`1996 I WB 342 I55
`I996]
`RM |6l.l49 I996
`6| S.3'3&—dc2O
`DNLMIDLC
`
`for Library of Congress
`
`96- 15357
`C IP
`
`The publisher offers discounts on this book when ordered in bulk
`quantities. For more information, write to Special Sales/’Professional Mar-
`keting at the address below.
`
`This book is printed on acid-free paper.
`
`Copyright © 1996 by Marcel Dekker, Inc. All Rights Reserved.
`
`Neither this book nor any part may be reproduced or transmitted in any
`form or by any means, electronic or mechanical, including photocopying,
`microfilming, and recording, or by any information storage and retrieval
`system. without permission in writing from the publisher.
`
`Marcel Deitker. Inc.
`
`270 Madison Avenue, New York, New York 10016
`
`Current printing (last digit):
`l0 9
`8
`7
`6
`5
`4
`3
`
`2
`
`I
`
`PRINTED IN THE UNITED STATES OF AMERICA
`
`2
`
`

`
`14
`
`Medical Devices for the Del lvery of Therapeutic
`Aerosols to the Lungs
`
`FIICHAHD N. DALBY
`
`and SUSAN L. TIAHO
`
`University all Maryland at Baltimore
`Baltimore. Maryland
`
`ANTHONY J. HIOKEY
`
`_
`_
`University or North Carolina
`Chapel Hi". NON“ Cflfflliflfl
`
`L
`
`Introduction
`
`Despite the numerous methods that can be employed to generate aerosols
`in therapeutically useful size ranges and concentrations (Chapters 9-1! and
`13), only three basic aerosol delivery systems have found their way into
`commercially marketed drug products. Specifically. these are metered-dose
`inhalers {MDIs), dry powder inhalers (DPIs), and nebulizers. These three
`classes of devices do not rep resent optimal delivery systems in terms of their
`ability to produce monoclispcrsc aerosols that can be precisely closed in a
`single breath. but rather are examples of delivery systems that achieve mini-
`mally acceptable characteristics in a simple. convenient, inexpensive. and
`portable format. To be acceptable for clinical use, an inhalation delivery
`system must meet certain criteria:
`
`1.
`
`2.
`3.
`
`It must generate an aerosol with most of the drug carrying parti-
`cles less than 10 pm in size. and ideally in the range 0.5-5 pm, the
`exact size depending on the intended application.
`It must produce reproducible drug dosing.
`It must protect the physical and chemical stability of the drug.
`
`44]
`
`3
`
`

`
`442
`
`,
`
`Dolby er of.
`
`4.
`5.
`
`It must be relatively portable and inconspicuous during use.
`It must be readily used by a patient with minimal training.
`
`These minimal requirements alone do not guarantee commercial suc-
`cess. Most commercial products currently under development aim to pro-
`vide multiple dosing (typically 200 doses] with minimal excipient inhalation
`(which can lead to poor organoleptic properties in the mouth. and oropha-
`ry:ngeal irritation). Patient convenience. competitive manufacturing costs
`to MDls. and added value features. such as dose counters or an indication
`of appropriate inhalation [low rates, are also considered desirable. In this
`chapter. the components. designs, and operating conditions of typical inha-
`lat.ion products are discussed. together with the possibilities that could be
`realized by “next generation” aerosol delivery systems.
`
`ll. Metered Dose Inhalers
`
`Since the 19505. MDls have been the mainstay of inhalation therapy. osten-
`sibly because they were perceived to meet most of the criteria outlined
`above. However. over the years a number of deficiencies have been identi-
`fied. Only a small fraction of the drug escaping the inhaler penetrates the
`patient's lungs (1.2) due to a combination of high particle exit velocity and
`poor coordination between actuation and inhalation. The unstable physical
`nature of suspended drug particles in propellant, combined with suboptimal
`valve designs, has led to reports of irreproducible dose metering following
`a period of rest (3). Low concentrations of potentially carcinogenic com-
`pounds were found to be extracted from valve components by the propel-
`lant system (4) and inhaled by the patient. However. the largest threat
`to the continued availability of pressurized MD!s is their dependence on
`chiloroiluorocarbon (CFC) propeliants, which have been linked to the
`depletion of stratospheric ozone and are now scheduled to be phased out
`under the terms of the “Montreal Protocol on Substances that Deplete the
`Ozone Layer" (5.6). Despite these concerns. new device designs. improved
`fotrrnulations and valves. and a switch to “environrnentally friendly" propel-
`lants are likely to keep the MD! in common use.
`The modern MD! shown in Figure I is little changed from its prede-
`cessors. and contains the same three basic ingredients: drug. one or more
`propellants. and in most cases. a surfactant. A liquefied propellant serves
`both as an energy source to expel the formulation from the valve in the
`form of rapidly evaporating droplets and as a dispersion medium for the
`drug and other excipieuts. A surfactant is typically present to aid with the
`dispersion of suspended drug particles or dissolution of a partially soluble
`drug. and to lubricate the metering value mechanism. In some formulations
`
`4
`
`

`
`Medical Devices for Aerosol Delivery
`
`443
`
`f
`
`:__p.
`
`fig ©-:::.
`
`C}
`
`Actuation
`
`Spraying or large
`drop-Iets moving at
`high velocity
`
`Evaporation
`
`Formation of small, slow-
`moving. inwvidual particles.
`aggregates. or partially
`evaporated droplets.
`
`Flgure 1 Diagram of a typical pressurized metered dose inhaler showing rnecha~
`nlsm of particle formation.
`
`a surfactant is reported to be unnecessary (7). Drug can be dissolved in the
`liquefied propellant./surfactant combination, with or without the aid of a
`less volatile cosolvent (8,9), or suspended in the form of rnicronized parti-
`cles (10). In all currently marketed formulations, drug dissolution necessi-
`tates the use of an ethanotic cosolvent. Flavors (such as dissolved mint
`extracts) and suspended sweeteners (for example, nticronized saccharine}
`may be present to combat the unpleasant taste associated with significant
`oropharyngeal deposition following inhalation. To enhance chemical stabil-
`ity, antioxidants (ascorbic acid) or chelating agents (EDTA) may be present
`in formulations in which the drug is dissolved.
`The popularity of traditional CFC propellants has stemmed from
`their low pulmonary toxicity, high chemical stability and purity, and com-
`patibility with commonly used packaging materials. In addition, they are
`nortflammable. Combinations of the three most widely used CFCs, trich|o-
`rofluorornethane (CFC-ll). dichlorocliliuoromethane {CFC~l2), and 1,2-
`dichlorotetrafluorontethane {CFC-I M), are typically combined in varying
`ratios to achieve a desirable combination of vapor pressure, liquid density,
`and solvency (1 1). Following a long search for alternative propellants with
`similar characteristics to CFCs. 1,1 ,l.2-tetrafluoroethane (HFC-l 343) has
`emerged as the primary replacement. and commercial formulations con-
`taining this propellant have recently gained or are awaiting marketing
`approval
`in several countries (12).
`In addition,
`l,I,l,2,3.3,3-hepta-
`fluoropropane (HFC—227) is being actively investigated. In the recent past,
`
`5
`
`

`
`444
`
`_;
`
`Dolby er at.
`
`numerous other propellants have been investigated ( I3}. and formulations
`containing them have been described in the patent literature {I4}. However.
`the expense of commercial development. primarily due to chronic toxicity
`testing, appears to have reduced the interest of many companies in these
`propellants. The physical and chemical properties of both CFCs and the
`emerging propellants are widely reported- Table 1 of Chapter 13 (p. 426)
`provides information on the important properties of the propellants men-
`tioned in this chapter. Gaseous propellants and nonpressurized sprays are
`not used due to decreasing internal pressures during use and the inability of
`pumps to produce small enough droplets. respectively.
`Historically. propellant blend compositions appear to have been
`somewhat arbitrarily selected. as evidenced by multiple drugs from the
`same company being delivered using the same propellant and surfactant
`combinations. This practice. while minimizing change over times during
`manufacturing. is unlikely to yield formulations that provide optimal drug
`delivery. The primary formulation characteristic dictated by the propellant
`is product vapor pressure. This determines the size and exit velocity (15) of
`the emitted spray and the rate at which propellant evaporation occurs.
`Careful optimization is essential. High vapor pressures produce faster exit
`velocities, which may lead to enhanced oropharyngeal deposition. How-
`ever, such blends also yield smaller, faster evaporating droplets, which
`facilitate production of smaller inhaled particles or droplets, which can
`enhance lung penetration. Conversely. a less volatile blend may minimize
`impaction in the throat at the expense of producing larger, slowly evaporat-
`ing droplets, which are prone to impaction high in the respiratory tract.
`This is apparent when the smaller particle size of an aerosol in which the
`compound is dissolved in pure CFCs is compared to one containing signifi-
`cant amounts of less volatile ethanol (9.10). The effect of vapor pressure
`on MD! leakage rates. transportation regulations. and filling process must
`also be considered.
`
`The propellant blend also dictates the product density since the other
`excipients are present at low concentrations. Large density differences be-
`tween the propellant blend and the true density of suspended drug particles
`are known to cause erratic dosing if there is a delay between shaking and
`actuation of the MD] (10). owing to the nonhomogeneous drug distribution
`within the canister due to drug sinking or floating in the propellant. This
`can be minimized by matching the drug and propellant density. or by facili-
`tating deflocculation of the bulk suspension with an appropriate surfactant.
`Propellant blending also allows the solubility of a drug or surfactant in the
`propellant system to be manipulated since some propellants (such as CFC-
`Il) are often better solvents than others {such as CFC-l2). However. it is
`important to remember that density and solvency cannot be manipulated
`
`6
`
`

`
`Medical‘ Devices for Aerosol Delivery
`
`445
`
`independently of vapor pressure. so in practice the changes that can be
`made are limited. One problem with the availability of only one alternative
`propellant (HFC-134a) is the loss of flexibility associated with the use of
`blends. Several papers and patents address this issue with the inclusion of
`ethanol as a cosolvent (I4). While this facilitates slurry production during
`manufacturing, reduces the vapor pressure. and enhances the solubility‘ of
`several surfactants in HFC-134a.
`it may result in a reduced fine particle
`fraction following MDI actuation.
`Oleie acid, soya-derived lecithin, and sorbitan trioleate have been
`widely used as surfactants in CFC-based formulations. The choice of con-
`centration is usually determined by experimentation in an attempt to max-
`imize the fine particle fraction and dosing reproducibility. Concentrations
`as high as 2% w/w have been used in concentrated suspension formula-
`tions, presumably to facilitate reliable valve operation in the face of an
`abrasive sprayed product. Since surfactants are nonvolatile, they increase
`the diameter of drug particles allowed to evaporate to dryness following
`aerosolization. and may also reduce the evaporation rate of propellant from
`sprayed droplets (8). Because of their unpleasant taste. and tendency to
`accumulate around spray orifices. excessively high concentrations are prob-
`ably best avoided. Much lower concentrations (0.01-0.1% vv/w) are often
`sufficient to facilitate homogeneous dispersion of suspended drug in pro-
`pellant following shaking. Less time-consuming, surrogate tests to identify
`appropriate surfactant concentrations. do not always correlate well with the
`observed fine particle fraction following spraying (16). All the surfactants
`listed above have been shown to exhibit maximum solubilities less than
`
`0.02% in Hf-‘C-134a, which has spurred a search for alternatives (17). The
`patent literature contains numerous references to potentially useful surfac-
`tants for use with HFC—l34a. These include polyethylene glycol, propoxy—
`Iated polyethylene giycols, perfluoroalkanoic acids, and numerous others,
`all of which exhibit enhanced solubility. Byron et al. have suggested that
`solubility may not be essential to the development of a successfully formu-
`lated suspension product if the drug particles are coated with an apparently
`insoluble surfactant that permits easy resuspension (17).
`Drug is either dissolved or suspended in a MDI formulation. The
`equilibrium size of a sprayed droplet containing dissolved drug depends on
`the starting droplet size and the concentration and density of the dissolved,
`nonvolatile ingredients it initially contained. In such a system, it is theoreti—
`cally possible to alter the drug concentration to achieve a wide range of
`sizes. If the droplet does not evaporate to dryness, which is likely if it
`contains at nonvolatile cosolvent. then the evaporation rate becomes the
`primary determinant of droplet size. in such formulations it is not necessary
`to reduce the size of the drug particles prior to incorporation into the
`
`7
`
`

`
`446i
`
`l"fl.|.
`
`Dolby er tn‘.
`
`formulation. Solution formulations are also easier to manufacture and do
`
`not. exhibit the drug-sedimentation-related problems described above. Be-
`cause of the intimate molecular interaction between the dissolved drug and
`excipients. reduced chemical stability compared to a suspension formula-
`tion has been observed. Thus. supersaturation and precipitation at low
`temperatures must be avoided. Partitioning of drug into valve elastomers
`has also been noted with a corresponding decrease in dose delivery per
`actuation (18).
`The aerosolized drug output from a suspension formulation cannot
`be smaller than the original particles used to prepare the suspension. Mi-
`cronization, spray drying, or controlled crystallization is therefore essential
`to produce appropriately sized particles. Gouda has shown that dilute sus-
`pension concentrations in the absence of dissolved ertcipients should theo-
`retically produce sprayed output of a similar size to the original drug parti-
`cles {l9). As the suspension concentration increases, multiple particle
`inclusion in the sprayed droplets leads to the production of aerosolized
`aggregates with a larger size than that of the starting drug. This effect
`becomes more significant at suspension concentrations greater than 1% v/ v
`and will also be exacerbated by the inclusion of other non volatile excipients.
`Experimental observations of model suspensions show the same trend {I9
`and references therein).
`In accordance with Stokes‘ law, smaller drug particles separate from
`the bulk suspension more slowly than larger ones. This offers a means of
`reducing the physical instability problems outlined previously if the suspen-
`sion cannot be adequately stabilized using surfactants. Suspension formula-
`tions can deliver larger doses than solutions since the forrnulator is not
`limited by the drug's maximum solubility in the propellant blend. However,
`high concentrations are likely to cause valve blockage and may abrade valve
`elastomers during filling (if the slurry is pumped through the value) or
`patient use. Partial solubility of a drug in propellant has been associated
`with an increase in the size of the suspended particles. Dalby et al. have
`reported a method of conveniently measuring drug solubility in pressurized
`systems (20), and Phillips ct al. (21) have documented its ability to predict
`crystal growth.
`As with other inhalation delivery systems, it is inappropriate to evalu-
`ate the formulation independently of the "packaging." The essential compo-
`nents are the container, metering valve, and actuator. Containers for sus-
`pensions are typically one-piece aluminum canisters, with a 20-mrn external
`crimp (cut edge or rolled top) neck. Some products (e.g., Azmacort) utilize
`more attractive epoxy-coated aluminum canisters. Some suppliers recom-
`mend anodized aluminum canisters for solution formulations. although
`most of these products are packaged in plastic-coated glass bottles. While
`
`8
`
`

`
`Medical Devices for Aerosol‘ Delivery
`
`44 7
`
`aluminum is inherently opaque, the coating on glass bottles is usually opaci-
`fied when used with light-sensitive drugs such as epinephrine. The container
`is typically required to withstand internal pressures up to 130 psig without
`distortion. The formulation type,
`labeling, aesthetic requirements. and
`need for in-process fill weight monitoring usually dictate the choice of
`container. Plastic containers are also available. but are not utilized in any
`marketed inhalation products.
`Metering valves (Fig. 2) are designed to release a fixed volume of
`product during each actuation. Assuming that the valve fills with a homoge-
`neous drug solution or suspension, the metered dose is the product of the
`valve volume and drug concentration. Usual valve volumes range from 25
`to 100 pl, although larger volumes are available. Typically. valves contain a
`prefilled metering chamber that is isolated from the bulk reservoir as the
`chamber empties through the valve stem. This is initiated by pressing the
`stern into the body of the valve. When pressure on the stem is released, an
`internal spring returns the stem to its rest position, and the metering cham-
`ber reflls through one or more channels from the reservoir. Such a valve is
`
`Skirt (uncrlmpad)
`
`Metering chamber
`filling orifice
`
`
`
`Valve seal
`
`Stem Side orifice
`
`Stern orifloe
`
`Flguro 2 Schematic diiagrarn of a metering valve.
`
`9
`
`

`
`448
`
`_;
`
`Dolby er in‘.
`
`described as “holding its prime," indicating that the product that will be
`sprayed next is already located in the metering chamber prior to actuation.
`and must be retained there if the next dose is to be complete. If drug escapes
`from the metering chamber during a period of quiescence (perhaps due to
`propellant drainage or suspension instability], the following actuation will
`release a smaller dose. Drug escape from the metering chamber of MDIS
`during periods of nonuse has been associated with a low subsequent dose of
`albuterol- For this reason. alternative valve designs are available. though
`no1'. without their own limitations (22). While drug concentrations present
`in the canister are altered by temperature-dependent changes in propellant
`density and propellant leakage, the past history of under- or overdosing by
`the valve should also be considered. Valve volumes are approximate due
`to propellant-induced swelling of the metering chamber elastomers and
`mechanical distortion during crimping and repeated use.
`It
`is therefore
`prudent to assess valve performance with specific formulations and not
`infer their reliability from successful use with a previous product. All meter-
`ing; valves on inhalation products are designed to operate in the "valve
`down” or “inverted" position and do not utilize a dip tube. This is because
`it is difficult to redisperse drug particles that sediment or float in a narrow
`tube. In addition, the valve must contain the pressurized product and retard
`the ingress of moisture or oxygen.
`Most valve bodies are constructed from plastics and resins, although
`metal body valves do exist. An aluminum fertile allows the valve skirt to be
`crimped on the canister. Return springs are usually stainless steel. Valves
`contain elastorneric seals in the metering chamber, and between the canister
`and the valve. Elastomer composition is frequently proprietary, and some-
`times unknown to even the valve manufacturer. Low concentrations of
`
`several potentially harmful chemicals are known to be extracted from valve
`elatstomers by propellant systems (4). This has led to the use of elastomer
`extraction procedures to reduce the concentration of these materials prior
`to valve assembly and in the development of “cleaner" elastomers. Elasto-
`mcr performance is critical to the functioning of a valve, since in combina-
`tion with other factors. it determines the propellant leakage rate. metering
`reproducibility. and the speed and reliability of stem return following actu-
`ation. Lirnited elastomer swelling may be considered beneficial since it helps
`en.sure a good seal. However. excessive swelling results in a nonfunctional
`valve. For this reason. the advent of new propellant systems has necessi-
`tated the development of new elastorners.
`The actuator is frequently the most visible part of the MDI. Its Func-
`tion is to make actuation of the valve easier. direct the spray into the
`patient‘s mouth. and provide the orifice through which the metering valve
`discharges its spray. A well-designed actuator with a separate or integral (to
`
`10
`
`10
`
`

`
`Medical Devices for Aerosol Delivery
`
`449
`
`prevent loss or accidental inhalation) dust cap protects the valve stem from
`damage and keeps it aligned with the seat. This prevents the accumulation
`and subsequent
`inhalation of dust, and provides a place for patient use
`information and product identity. The spray orifice and valve stem seat are
`arguably the most critical parts of the actuator. Larger spray orifices are
`often used in combination with large volume metering valves to spray con-
`centrated suspensions with a minimized likelihood of blockage. Large ori-
`fices ensure fast emptying. When small volumes of dilute solutions and
`suspensions are sprayed, smaller spray orifice diameters may be preferred
`since they generate smaller droplets (15). To avoid leakage. a tight fit
`between valve stem and actuator seat is essential. Additionally. it is essential
`to minimize sharp turns and blind ends in the path the product follows
`from the valve stem to the spray orifice to prevent accumulation. and
`subsequent blockage. by nonvolatile drug and eltcipients.
`With the exception of Rhone-Poulenc Rorer’s Azmacort(Collegevi1le,
`PA). and Astra’s Breathancer (Luna, Sweden}. all MDls are supplied with
`a molded plastic actuator, which positions the patient‘s lips very close to the
`spray orifice (if they use the closed-mouth method of inhaler use). This
`provides a short distance between the spray orifice and oropharynil and
`necessitates excellent coordination between actuation of the MDI and inha-
`
`lation by the patient if almost complete oropharyngeal deposition is to be
`avoided. Spacer devices (Fig. 3) were developed to increase this distance.
`allowing the rapidly advancing aerosol cloud to decelerate before reaching
`the throat (11). This makes perfect synchronization between actuation and
`inhalation slightly less important. In addition. spacers allow more time for
`
`
`
`Figure 3 Schematic diagram of a metered dose inhaler and reservoir device (Nebu-
`haier).
`
`11
`
`11
`
`

`
`450
`
`,
`
`Darby er cl.
`
`propellant evaporation. resulting in the formation of smaller droplets or
`particles and less reflex coughing and exhalation due to local cooling of the
`throat by impacted. evaporating droplets. A large proportion of drug that
`would otherwise deposit in the orc-pharynx is retained in a spacer. This
`reduces systemic drug levels and minimizes local side effects. A distinction
`between spacers and holding chambers {also called reservoirs) is increas-
`ingly drawn. Spacers are essentially hollow tubes through which a patient
`should have started inhaling prior to actuating the MDI. They are designed
`to empty in a single inhalation. Reservoirs are typically larger in diameter.
`frequently conical or pear—shaped devices. and are designed to permit actua-
`tion of the MD! prior to initiating an inhalation. Their larger size is de-
`signed to reduce drug losses on the interior wall of the reservoir due to
`impaction and sedimentation. Reservoirs usually contain a one-way valve
`to prevent an inadvertent exhalation from flushing a previously aerosolized
`dose from the device. Delays between actuation and inhalation. making
`multiple actuations in the reservoir. and emptying the reservoir over several
`inhalations all reduce the efficiency of aerosols delivery to the lung.
`Because larger spacers and reservoirs more effectively enhance lung
`delivery compared to smaller ones. collapsible designs are common (e.g..
`Inspirease. Schering Corporation, Kenilworth. NJ}. Other designs direct
`the emerging aerosol spray in the opposite direction to the inhaled airstrearn
`in an attempt to increase the flight time while minimizing device size (Opti-
`Haler, Healthsean Products. Cedar Grove. NJ). Spacers and reservoirs may
`also contain flow restrictors to control the patients inhalation rate and have
`mechanisms to coordinate inhalation with MD! actuation (Optil-laler).
`Many audibly warn the patient when they are inhaling too fast [Aerosol
`Cloud Enhancer (ACE). DHD Die-molding l-Iealthcare Division, Canas-
`tota, NY: AeroChamber, Monoghan Medical Corporation, Plattsburgh,
`NY; Inspircasc]. Newer devices are typically transparent to encourage regu-
`lar cleaning and some are designed to fit into ventilator circuits. Baffles
`located within several small actuators have been shown to yield many of the
`same advantages (23). Because spacers and reservoirs are often designed to
`ill multiple MDIs (which may have significantly different compositions.
`valves, and actuators), their ability to equally enhance the delivery of all
`products has been questioned.
`The Autohaler (3M Pharmaceuticals, St. iiaul. MN) is a small device
`that uses a mechanical vane to detect when a patient's inhalation rate is
`appropriate for automatically firing the proprietary MD! it contains. While
`achieving excellent coordination between inhalation and actuation, it does
`not produce the other advantages associated with a spacer or reservoir.
`MDIS from other manufacturers cannot be used in the Autohaler. Elec-
`
`tronic devices capable of more sophisticated flow monitoring by program-
`
`12
`
`12
`
`

`
`Medical Devr'ces' for Aerosol’ Delivery
`
`451
`
`ming actuation at different flow rates or points in the breathing cycle, and
`which can record the history of patient compliance. are under development.
`
`Ill. Dry Powder Inhalers
`
`Dry powder inhalers offer a unique opportunity for the delivery of drugs to
`the lung as aerosols. These devices combine powder technology with device
`design in order to disperse dry particles as an aerosol in the patient’s inspira-
`tory airflow (24). Powders have been insufflatecl for medical purposes
`throughout history (25). It is only recently that efforts have been made to
`establish the dispersion properties of particles and their impact on therapeu-
`tic effect (26). A great deal of progress has been made in recent years as the
`emphasis has changed from unit dose systems employing only the patient’s
`breath to generate the aerosol, to multiple-dosing reservoir devices that
`actively impart energy to the powder bed to introduce drug particles into
`the inspiratory airflow.
`All DPIS have four basic features: (1) a dose-metering mechanism,
`(2) an aerosolization mechanism, (3) a deaggregation mechanism, and (4)
`an adaptor to direct the aerosol into a patient’s mouth. The major compo-
`nents of a dry powder inhaler are the drug powder, and other powdered
`excipients where necessary. a drug reservoir or premetered individual doses,
`the body of the device. and a cover to prevent ingress of dust or moisture.
`To introduce drug particles into the lung, they must be (5 pm in
`aerodynamic diameter (27,28)- This is generally achieved by milling the
`powder prior to formulation (29,30). In recent years there has also been
`some interest in spray drying powders to achieve the same end (31-33).
`Small particles are notoriously difficult to disperse (34). The forces govern-
`ing dispersion are well documented and consist mainly of electrostatic, Van
`der Waals. and capillary forces (35). Knowing that these forces exist has
`not facilitated aerosol generation to any great extent. One approach that
`has been taken to improve the dispersion of dry powders is the inclusion of
`an excipient, notably lactose (36,37). The lactose particles are intended to
`act as carrier particles for the drug and as such are in a much larger size
`range, 60-80 pm (38). Drug particles are theoretically stripped from the
`surface of the lactose panlcles, to which they are loosely attached, during
`the generation process (39). This process is illustrated schematically in Fig»
`ure 4. Thus, the drug particles are dispersed and can traverse the upper
`respiratory tract while the excipient particles do not pass beyond the mouth-
`piece of the device or the mouth and throat of the patient.
`In the devices that have been approved for use in the United States,
`the Spinhaler (Fisons, Rochester, NY) and Rotahaler (cilaxowellcorne.
`
`13
`
`13
`
`

`
`452
`
`Dolby e: at‘.
`
`
`
`.
`
`h
`
`B U
`O '
`.
`"
`.
`-'*'-D
`('3 u
`.
`,
`0 O . O ‘I
`I
`il 3
`
`airflow
`—-I 7.‘?
`
`6
`
`eamoridrttqetalio
`powder bed
`
`earrierandstmoeddrttu
`some-ol dispersion
`
`Figure 4 Schematic diagram of the stripping of drug particles from carrier parti-
`cles on the inhaled airstream.
`
`Inc., Research Triangle Park, NC) unit doses of Intal (disodium cromogly-
`cate) and Ventolin (albuterol sulfate}, respectively. are packaged in hard
`gelatin capsules. The dose of powder itself is delivered from the gelatin
`capsule by different mechanisms. The Spinhaler has a mechanism for pierc-
`ing the capsule (40). The cap of the capsule fits into an impeller. which
`rotates as the patient breathes through the device projecting particles into
`the airstream, as illustrated in Figure 5. The Rotahaler. shown in Figure 6,
`
`rotating impeller
`
`aerosol pattides
`
`Figure 5 Diagram indicating the essential components of at Spinhaler.
`
`14
`
`14
`
`

`
`Medical Devices for Aerosol Delivery
`
`453
`
`has a mechanism for breaking the capsule in two pieces. The capsule body
`containing the dose falls into the device. while the cap is retained in the
`entry port for subsequent disposal. As the patient inhales. the portion of
`the capsule containing the drug experiences erratic motion in the airstrearn.
`causing dislodged particles to be entrained and subsequently inhaled.
`A number of other devices are approved for use in other countries.
`The lnhalator (Boehringer Ingelheim, Ridgefield, CT) has a mechanism for
`piercing the ends of a hard gelation capsule containing the dose of fenoterol
`(41), as shown in Figure 7. The inspiratory flow then passes through the
`capsule. The Dislthaler (filaxowellcome. Inc., Research Triangle Park,
`NC), shown in Figure 8, employs packaging consisting of individual doses
`of albuterol sulfate in blister packs on a disk cassette. Following piercing,
`inspiratory flow through the packaging depression containing the drug in-
`duces dispersion of the powder. One of the more sophisticated systems
`approved for use in Europe is the Turbuhaler (Astra Pharmaceuticals,
`Lund, Sweden). shown in Figure 9 (42). This device employs a multidose
`reservoir of terbutaline sulfate. The dose is metered into small conical cavi-
`
`ties by twisting a grip at the base of the device. When the patient inhaies.
`air ducted through the cavities dislodges a dose of drug.
`In addition to the devices mentioned above, many others patented for
`use are described in the literature (though none have so far received regula-
`tory approval in the United States). However. the novelty of these systems
`is typically associated with their mechanism for aerosolizing the powder
`rather than the way a unit dose is packaged or separated from a bulk
`
`
`
`ENDS DI
`
`Figure 6 Diagram indicating the essential components of a Rotahaler.
`
`15
`
`15
`
`

`
`454
`
`Dafby 91‘ al.
`
`
`
`Figure 7 Diagram indicating the essential components of an lnhalator.
`
` sup-pun disc
`
`./
`
`mouthpiece cover
`
`mouth piece orifice
`
`Figure 8 Diagram indicating the essential components of a Diskhaler.
`
`16
`
`16
`
`

`
`Medical Devices for Aerosol Delivery
`
`455
`
`
`
`Figure 9 Diagram indicating the essential components of a Turbuhaler.
`
`reservoir. Reproducible dose metering remains the most difficult challenge
`in device design. For this reason, prefillcd closes in gelatin capsules or
`multiple depression blister packages. or extraction of a specified volume of
`free flowing powder from a multidose reservoir. are likely to remain com-
`mon. The volumetric metering of powder poses some unique problems since
`p