Drug Delivery Devices
`Issues in Drug Development
`
`Peter R. Byron
`
`Virginia Commonwealth University, Richmond, Virginia
`
`New techniques for delivery of drugs by inhalation are discussed
`in this article. Devices that promise to improve the efficiency of lung
`delivery are described along with some of the regulatory challenges
`faced by their development scientists. Although high delivery effi-
`ciencies are possible, such devices are expensive to develop and
`may only be feasible in the event that they are partnered with
`drugs whose therapeutic and economic value is truly enhanced by
`the effort invested in the process. Appropriate devices must also
`be selected after paying careful attention to the physicochemical
`and dosing demands associated with the drug substance to be.
`inhaled. Even newly launched commercial products display large
`variations in dose delivery to the lung, in spite of increased global
`efforts to regulate and ensure the uniformity of delivered doses
`and their aerosol size distributions; this because of variations in the
`inspiratory maneuvers used by patients and the lack of control
`exercised over these maneuvers by most new inhalers. Sophisti—
`cated electromechanical techniques are discussed as possible ways
`of overcoming some of the common difficulties associated with
`ensuring reproducibility of dose and drug delivery to the lung.
`
`Keywords: aerosol; drug; delivery devices; inhalers; lungs
`
`After a long period during which chlorofluorocarbon (CFC)-
`pressurized metered dose inhalers (MDIs) dominated the mar-
`ket, in the last 15 years several new inhalation devices have been
`introduced. Other novel devices remain “in the wings,” and these
`may yet be exploited by pharmaceutical companies for specific
`reasons (1). We are seeing increasing acceptance of the lungs
`as a route of systemic drug administration, as well as a number
`of new drugs offering novel therapeutic benefits for several acute
`and chronic lung disorders (2, 3). These developments represent
`significant opportunities for pharmaceutical companies, pro-
`vided they choose delivery systems that adequately “partner”
`each drug during its development. The regulatory hurdles facing
`inhaler developers have become more stringent as the global
`marketplace has extended the impact of the U.S.A.’s Food and
`Drug Administration; FDA is probably the strictest of the world’s
`regulatory authorities. Multidose inhalers must now be shown
`to deliver individual doses reproducibly throughout inhaler shelf—
`life, in temperatures and humidities that represent commonly
`experienced environmental conditions. Not only must doses be
`reproducible, but the particle or droplet size distributions from
`each inhaler must also be shown to be “stable” over the product’s
`lifetime, and the product proven to be manufactured reproduci-
`bly (3). This must be done to show that the clinical results that
`are presented to the regulators at the time of product submission
`are “representative” of the “to be marketed” inhaler. Some of
`the recent increased regulatory oversight has been stimulated
`by environmental demands to replace the CFC propellants in
`
`(Received in original form March 5, 2004; accepted in final form lune 76, 2004)
`Correspondence and requests for reprints should be addressed to Peter R. Byron,
`Ph.D., Professor and Chairman, Department of Pharmaceutics, Wrginia Common-
`wealth University, Richmond, VA 23298—0533. E-maiI: prbyron@vcu.edu
`Proc Am Thorac Soc Vol 1. pp 321-328, 2004
`DOI: 10.1513/pats.200403-023MS
`Internet address: www.atsjournals.org
`
`preSSUIiZed MDIs with non-ozone—depleting hydrofluoroalkanes
`(HFAs), at a time when many older drugs and inhalers were
`becoming subject to competition from the generic industry (1).
`Not surprisingly, there has been considerable commercial incen~
`tive, especially for innovator companies wishing to market new
`drugs, to attempt to launch such drugs in novel inhalers.
`From the drug point of View, inhalers contain mostly short-
`or long~acting topical bronchodilators (adrenergics as well as
`anticholinergics), antiinflammatory steroids, and antiallergics
`such as cromolyn sodiumr(3). We are also seeing proteins being
`delivered by this route for their loeal actions. Drugs like a—l-anti—
`trypsin and other antiproteases, genes, and oligonucleotides are
`all in clinical trial for local effects in the lung. Nebulized thNase
`was launched by Genentech some 5 years ago for treatment of
`patients with cystic fibrosis. Antiviral agents and vaccines are
`also under development for delivery as aerosols to enhance
`“local immunity” in different parts of the respiratory tract (2—6).
`In the systemic drug area, only ergotamine has been traditionally
`delivered through the lung following its presentation as a me-
`tered dose (pressurized) inhaler (3). However, biotech compa-
`nies are attempting to deliver a number of hormones systemically
`via the lungs (e.g. calcitonin, hGH, PTH), and insulin is presently
`the subject of large Phase III trials by both Aventis—Pfizer and
`Novo~N0rdisk with their respective inhalation partners Nektar
`Therapeutics and Aradigrn Corporation (5, 7). In summary, the
`safety of the pulmonary route has gradually gained acceptance
`for macromolecular delivery (8) and there is a tremendous
`amount of interest and drug company research activity in this
`area. This will result in a large number of changes to inhaled
`drug therapy in the future.
`
`IMPROVING DRUG DELIVERY EFFICIENCY
`TO THE LUNG
`
`From the point of View of those wishing to pursue inhaled drug
`development, there are a number of frequent questions that are
`asked. These are presented, along with some rather general
`answers, in Table 1. It is clear from this Table that the choice
`of drug delivery technology to be pursued requires some unique
`expertise. “Efficient” dose delivery to the lung can be discussed
`by defining efficiency as the percentage of the dose depositing
`in the airways (DTL) relative to the delivered dose (dd) or, in
`some cases, the label claim on the inhaler (Table 1; Efficiency =
`100 - DTL/dd). It should come as no surprise to learn that high
`efficiencies are possible but, of course, the price of moving in
`that direction can be considerable. Motivating companies to
`accomplish high delivery efficiencies (in contrast to achieving
`reproducible but inefficient lung penetration) requires that some
`penalties are associated with less efficient delivery. In many
`cases, the lack of a penalty means that these “motivations” do
`not always exist. However, examples may be related to toxicol-
`ogy (e.g., oral corticosteroid deposition associated with candidia-
`sis and unnecessarily large total drug exposure), therapeutic
`need (e.g., insulin and other hormones require peripheral lung
`penetration to enable reasonable bioavailabilities to be achieved;
`3, 5), product misuse by patients (e.g., actual use of product
`encourages additional dosing and an increased incidence of ad-
`
`WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 1 of 8
`
`WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 1 of 8
`
`

`

`322
`
`PROCEEDINGS OF THE AMERICAN THORACIC SOCIETY VOL I
`
`2004
`
`TABLE 1. FAQS CONCERNING INHALER DEVICES WITH ANSWERS RELEVANT
`TO DEVELOPMENT SCIENTISTS
`
`Can N 100% of each "metered dose” be delivered to lung? No, all devices retain some of the (premetered) drug substance,
`and much of what is delivered is often deposited in the oropharynx.
`Can a large proportion of the "ex-mouthpiece dose” be delivered to lung? Yes, the ex-mouthpiece dose, or delivered dose‘,
`can be made to emphasize only the small "more respirable” particles.
`Can the close delivered to the lung (DTL’) be directed preferentially to the peripheral and/or the central ainivays? Yes, this
`is possible for normal individuals but it is not easily achieved for those with lung diseases.
`Are large payloads possible? Yes, values forDTL perinhalation S 4mg are possible. Above this value, multiple breaths or continuous
`inhalation (e.g. nebulizers) must be employed.
`Can “fragile molecules” be delivered intact? Yes, although process and formulation constraints are drug-specific.
`Are 2 2—year product shelf-lives possible? Yes, but formulation packaging and the use of solid-state drug in formulations are
`extremely important features in cases where reproducibility is a problem.
`Can the delivered dose, dd, and its emitted particle size distribution, psd, be made reproducible throughout inhaler lifetime?
`Yes. If this is not so, the product will not be accepted by regulators.
`Are all devices suitable for all drugs? No, the different device platforms are not generally applicable.
`Are some devices easier to develop? Yes, but present/y these are not the most efficient. High delivery efficiency platforms are
`more expensive to develop and launch.
`
`* Delivered dose, dd, is the dose leaving the inhaler mouthpiece when the inhaler is used according to the package insert.
`T The dose leaving the inhaler that deposits in the airways.
`
`verse events; 3, 4), or drug cost (e.g., drug substance costs consti-
`tute a large and prohibitive part of each inhaler’s value; 3).
`To achieve highly efficient drug delivery to lung requires that
`the aerosol creation process be controlled so that a fine droplet
`or particle cloud, containing a known drug concentration in a
`defined size distribution, is metered reproducibly into the pa—
`tient’s inhaled air stream. Provided a normal individual is inhal-
`ing slowly (o3 0.5 L/second), aerosol sizes, expressed as aero-
`dynamic diameters, should preferably be 3 to 7 pm for good
`tracheobronchial deposition, or between 1 and 3 pm for deposi-
`tion in the pulmonary regions (6). Breath holding is known to
`further enhance deposition in the pulmonary or alveolar regions
`(6). Unfortunately, although device designers can often accom~
`plish some of these requirements, patients rarely comply well
`with instruction leaflets given with inhalers. As a result, they
`often require regular counseling to reduce their use of inappro-
`priate inspiratory maneuvers, and in particular, often have diffi-
`culty coordinating their inspiration with the creation of the aero-
`sol cloud by the device (3, 4). For this reason, the reproducibility
`of DTL in vivo is often much worse than the apparent reproduc-
`ibility of the delivered dose assured the regulators during the
`drug development process (DTL and dd have large and small
`variance, respectively; Table 1). Furthermore, the presence of
`acute or chronic obstructive airways disease(s) in patients often
`precludes the delivery of a significant proportion of each dose
`to the lung periphery, irrespective of whether a device is selected
`which enables the predominant production of small aerosol par-
`ticles (1—3 pm) theoretically capable of penetrating the region
`in normal healthy volunteers (6).
`
`(9). There are two major types of hand—held nebulizer (3, 4).
`Air blast or “jet” nebulizers are cheapest and most likely to be
`used clinically, whereas continuous—output ultrasonic nebulizers
`have fallen out of favor recently for several reasons, including
`their inability to nebulize suspension formulations (10). Because
`the pharmaceutical vehicle used in nebulizer formulations is
`essentially aqueous, patients can use nebulizers for prolonged
`periods of time, to deliver quite large doses to the lung; tens of
`milligrams as opposed to micrograms are feasible in many cases.
`It is important to recognize that jet nebulizers can work quite
`well to create less than 5 pm, highly respirable, droplet aerosols
`for delivery of compounds that are either in solution or formu-
`lated as microfine—suspensions and placed in nebulizer reservoirs.
`There is thus no technologic reason why nebulizer therapy
`should not be used for effective drug delivery to the lung. There
`are, of course, economic reasons why they may be less than ideal
`if people are forced to use poorly designed, cheaply manufac~
`tured (variable) nebulizers repeatedly, as has been the case in
`some health centers. One problem then is economic; another
`may be the issue of convenience, because nebulizer systems are
`usually not very portable due to the need for a compressed gas
`supply.
`Most jet nebulizer designs force pressurized gas (usually air)
`from a nozzle (or jet), at high velocity past a liquid feed tube,
`so that the nebulizer solution is atomized at the capillary exit.
`The bulk of the aerosol mist (which may be traveling at up to
`sonic velocity) impacts against a baffle, drains back into the
`reservoir in the base, and recirculates. Only very small droplets
`(< N 5 pm, if the system is running correctly) escape the baffle
`and are available for inhalation. These droplets are produced
`at device-specific airflow rates and are inhaled along with “dilu-
`tion air” through a mouthpiece or mask arrangement. Clearly,
`Thus, the major factors that defeat inhaler design intentions are
`patient variables. Clearly, in the case of nebulizers used with
`some 50% of any aerosol produced by a nebulizer that is continu-
`sedentary patients, especially inhalers designed to deliver me-
`ously producing aerosol cannot be available during the patient’s
`exhalation because they are breathing tidally. Moreover, jet and
`tered doses, the inspiratory flow, tidal volume, and frequency
`baffle designs of different systems have hugely different efficien-
`of breathing are defined by the gas exchange needs of the patient
`cies, resulting in large differences between the times taken to
`and we may consider that some “control” exists over the patient’s
`administer a similar DTL, even from the same solution formula-
`average breathing pattern. If a nebulized aerosol is then pro-
`duced in a continuous fashion, patients may inhale and exhale
`tion. In all cases, aerosol generation quits before the entire drug
`its ouput for several minutes while seated. Unfortunately, the
`has left the reservoir. Most recently, several nebulizer manufac-
`way in which drugs are marketed as nebulizer solutions means
`turers (e.g., Pari, Aerogen) have designed systems to minimize
`that some of the potential “control” offered by this situation is
`‘drug losses during patient exhalation and minimize the times
`usually discarded by companies who market solutions for use
`taken to generate and deposit a given DTL (9). Thus, over the
`with undefined devices, using undefined operating conditions
`last 10 years these devices have become more efficient delivery
`WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 2 of 8
`
`NEBULIZERS
`
`WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 2 of 8
`
`

`

`i
`
`——‘.-::==:..
`
`
`.M.__.¢.Imo-
`
`systems, but only if the physician insists that patients use the
`most modern technology. In general, if we choose reputable
`devices, define the amount of drug placed into the reservoir in
`a uniform volume (e.g., 3 ml), and ask patients to inhale and
`exhale through them until the system quits, then DTL should
`be somewhere between 8 and 14% of that which was initially
`instilled into the reservoir. The remainder is lost during exhala-
`tion or it is retained in the device. Aerogen, Omron, and Pari
`market hand-held ultrasonic (piezoelectric) devices that are
`smaller and less cumbersome than the compressor—driven sys-
`tems described above. These generate slightly larger aerosols
`and can accomplish similar values for DTL more quickly than
`air-blast systems. Although overall efficiencies are likely to con—
`tinue to improve for high-cost systems (perhaps approaching
`20—25% of the instilled dose), the overall efficiency of the present
`piezoelectric systems is frequently found to be similar to that
`for jet nebulizers (3, 4). Most recently, some device manufactur-
`ers have brought to market computer-controlled systems that
`monitor each patient’s breathing pattern and administer nebu-
`lizer output phased with inspiration, whence to enable control
`over DTL (11).
`From a clinical perspective, drugs in nebulizer solutions may
`be deposited in larger doses than those seen with many metered
`delivery devices. Also, patients who are seriously afflicted with
`obstructive lung conditions prefer to use nebulizer therapy, be-
`cause of the nebulizer’s generally more “respirable” output, its
`continuous generation, and its aqueous, often buffered, vehicle
`(12). However, assuming that the drug is an appropriate choice,
`nebulizer therapy can still fail to be effective because of poor
`nebulizer or compressor selection and an unsuitable choice of
`operating conditions. Notably, with more recent drug launches,
`Genentech’s “Pulmozyme” package insert is instructive. In ef-
`fect, this recommends the use of a small number of specific
`nebulizer systems with this product; only those that were found to
`be equivalent for in vitro/thNase delivery during Pulmozyme’s
`development are recommended in the package insert. This “co—
`marketing approach” is likely to occur more frequently for newly
`launched nebulizer drugs. Finally, it is important that nebulizer
`solutions should not be stored in nebulizer reservoirs; stored
`
`solutions may grow bacteria which can then be nebulized and
`inhaled. Devices should be washed and dried between uses, and
`dishwasher compatibility may thus be important.
`
`DOSE-METERING INHALER SYSTEMS
`
`Pressurized metered dose inhalers, dry powder inhalers, or other
`“bolus aerosol”-producing technologies are considered to be
`“controlled dose” or “dose-metering” systems because they are
`sold in forms in which they can only be used with a single drug
`and formulation. We should appreciate, however, that education
`and compliance continue to be issues that need to be addressed
`with all devices, if we hope to be able to control DTL (3, 4, 7).
`Metered aerosols are designed to be inhaled as boluses, unlike
`the homogeneous aerosols distributed throughout inhaled air by
`nebulizers. Bolus aerosols are known to deposit in different
`places, dependent upon when, in the inhalation, the “bolus” is
`released. In addition, many bolus pharmaceutical aerosols are
`unstable physically; that is, they may evaporate and/or have high
`tendencies to impact, decelerate, and/or sediment over short
`time periods.
`From a drug development perspective, there are numerous
`formulation issues that can impact the reproducibility of deliv-
`ered doses and aerosol size distributions from dose-metering in-
`haler systems. A thorough review of this area is beyond the scope
`of this article. However, the subject has been discussed elsewhere
`by the author and others from the point of view of pressurized
`
`metered dose inhalers (1, 13), powder inhalers (3, 14—18), and
`the physwal difficulties that are often found when attempting
`reproducible metering with extremely potent and unstable com—
`pounds, like formoterol (19).
`
`Pressurized MDIs and Accessories
`
`MDIs remain the “gold standard” delivery system in many senses
`even though considerable efforts have had to be made to rede-
`velop acceptable valve and packaging systems, and to reformu—
`late products with HFA propellants (1). HFA-propelled MDIs
`are easily portable, tamper—proof, and multidose. They protect
`the remaining pressurized, liquefied product from oxidation,
`light, and water ingress while providing an inexpensive, mature
`technology with accurate liquid dose metering by volume (1, 3).
`Although reformulated systems have often chosen to mimic the
`inefficiency of the older CFC~propelled formulations, solution
`formulations are often feasible which can achieve values of DTL
`that approach half of the metered dose and exceed 50% of the
`delivered dose (20); this provided that close requirements are of
`the order of 100 ug or less. Moreover, MDIs are apparently
`easy to use, even though a considerable literature exists to show
`that in vivo values for DTL are variable, because of the difficul-
`ties that patients experience trying to coordinate inhalation with
`actuation of the MDI (4). The aerosol drug dose exits the MDI
`mouthpiece as a rapidly moving large droplet cloud. However,
`at about the distance of the back of the throat, the droplet
`diameter is reduced due to propellant evaporation, and a reason-
`able proportion of the “polydispersed” aerosol cloud is now
`small enough to penetrate the lung. It should not be surprising,
`therefore, that a proportion of each “metered dose” is lost in
`the actuator mouthpiece, and a further proportion is lost in the
`oropharynx due to inertial impaction of the “ballistic portion”
`of the spray. Spacers and reservoirs have come onto the market
`as compliance aids. These devices are shown diagrammatically
`in Figure 1; reservoirs can be differentiated from spacers because
`they contain a valve of some description, intended to retain the
`aerosol cloud created by the MDI, until the patient inhales.
`Spacers, on the other hand, contain no such value and simply
`distance the inhaler mouthpiece from the patient’s oropharynx.
`Both devices can reduce drug deposition in the back of the
`patient’s throat and enable extra time for evaporation to occur
`(4). Waste drug that would have been captured at the back of the
`patient’s throat can be partly retained in the spacer or reservoir.
`Notably, there is no reason why MDIs should not be developed
`with an attached spacer or reservoir for just this purpose, al—
`though many of the claims for the advantages of these devices
`fail to exemplify the problems known to be associated with
`multiple actuations into reservoirs, untoward delays between
`cloud generation and inhalation, and the effects of spacer and
`reservoir electrostatic charges; all of these effects dramatically
`reduce the DTL per dose from the MDI (13, 21). Notably, MDIs
`were rarely tested for efficacy or safety with such devices attached
`and their addition could therefore be construed by drug regula-
`tors as an example of product misuse (3, 4).
`Probably the greatest improvements likely to be associated
`with MDI usage in the future will involve the marketing of these
`devices as breath-actuated inhalers. There are many such new
`devices becoming available at present, although 3M has been
`marketing Autohaler (a breath~actuated CFC MDI) for a num—
`ber of years. Figure 1 and its legend describe an expensive option
`called Smartmist, developed by Aradigm Corporation and regis~
`tered (but presently not marketed) in the United States, with a
`number of excellent and sophisticated features (11). The device
`accepts and seals around regular pressurized MDIs and subjects
`these inhalers to microprocessor control. A built-in spirometer
`ensures that the patient receives the drug at the correct point
`
`WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 3 of 8
`
`WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 3 of 8
`
`

`

`324
`
`PROCEEDINGS OF THE AMERICAN THORACIC SOCIETY VOL I
`
`2004
`
`Lion
`
`Evapora
`, S
`
`imwnfl
`
`fitmme mm m if,
`
`WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 4 of 8
`
`WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 4 of 8
`
`
`

`

`‘3
`
`25
`
`These either measure the dose themselves (from a powder reser—
`voir) or they dispense and disperse individual doses which are pre-
`metered into blisters by the manufacturer. Turbohaler and Diskus,
`from AstraZeneca and GlaxoSmithKline, respectively, are rep-
`resentatives of the former and latter categories, although many
`other different designs are presently in development (17, 18, 23).
`In general, the premetered drug-in-blister approach is the easier
`type to develop because the reproducibility of the metered dose
`can be assured during the drug formulation packaging and device
`assembly processes. Provided then, that the device and formula-
`tion design is such that (1 ) the blisters empty well and (2) formula-
`tion and drug adhesion to the device is minimal, the delivered dose
`can usually be shown to the regulators to be reproducible (17, 18).
`With a device like Turbohaler, however, whose operating princi-
`ples are well known (23) and will not be repeated here, the
`powder metering system delivers doses in vitro that are much
`more variable than those derived from premetered blister packs.
`This intrinsic variability, which increases after storage and/or
`transport, must therefore be shown in the clinic (for each sepa-
`rate drug formulation) to have no therapeutic, or toxicologic
`consequences. The effort is complicated further if we observe
`that even with quite sophisticated powder inhaler devices, the
`mode of use and ambient environmental conditions can often
`
`define drug doses reaching the patient’s lungs (15 , 16). The advice
`to “keep your powder dry” continues to hold, and in many of
`these devices, designers and manufacturers have gone to some
`lengths to reduce the likelihood of water vapor ingress creating
`problems. With Turbohaler and others, desiccant is included
`(23) and the overcap should be kept firmly in place when the
`inhaler is not being used. Exhalation into the device, which
`should never be practiced, is also often addressed in patient
`information leaflets.
`
`The systems discussed so far are known as “passive” DPIs,
`because of their reliance upon “patient power” for the purposes
`of drug aerosolization. These all deliver variable doses and show
`delivery efficiencies to the lung which depend on the inspiratory
`effort the patient expends during inhaler use (inhale faster
`through these and DTL, even the delivered dose in some in-
`stances, increases). Ironically, however, clinical studies with Tur-
`bohaler (which shows high in vitro variability; 15) show that its
`in vivo dosing variance is smaller than that of the MDI (24),
`which suffers from high in viva dosing variance, due to coordina-
`tion problems at the patient interface. In vitro, the delivered
`dose can also often be shown to be dependent on a term known
`as “acceleration” or the rate at which a given airflow through
`the inhaler is approached during testing (17). We have illustrated
`
`Byron: Drug Delivery Devices
`
`in the inspiration, (e.g., shortly after they commence inhalation
`from residual volume). It also ensures that they are inhaling
`slowly as fast inhalation will require repeat attempts; an indicator
`light will inform the patient if the dose is inhaled (subsequently)
`at the correct rate. Because the device’s memory can be down-
`loaded, the health professional can check whether or not people
`are complying with recommendations. This type of technology,
`therefore, is clearly one way of removing some of the control
`from the patient, and thus ensuring improved benefits by virtue
`of reducing the variance of in vivo DTL (4, 22); at the very least,
`it should enable improvements to be made in the analysis of
`results during and following ambulatory clinical trials.
`
`Dry Powder Inhalers
`
`The present popularity of novel dry powder inhaler (DPI) devel-
`opment commenced at the time that CFC replacement became
`an issue. These devices are now much more sophisticated than
`used to be the case when only single-dose, capsule—loading sys-
`tems existed (e.g., Spinhaler and Rotahaler from Aventis and
`GlaxoSmithKline, respectively). From a user perspective, the
`number of different designs that are marketed or in development
`will itself create problems for physicians and patient educators.
`We already have difficulty teaching patients to use MDIs; imag-
`ine the increasing complexity as the variety of inhaler options
`increases further. lmportantly, from the point of view of DTL
`variance, most DPIs only deliver drugs when the patient inhales
`through them. As a consequence, the issue of “coordination”
`between actuation and inhaling disappears. However, because
`small volume powder metering is never as precise as the measure-
`ment of liquids, and because DPIs are generally less robust than
`MDls, there are many alternative (and often inhaler—specific)
`ways in which patients can misuse these inhalers. As a result,
`patients may fail to receive therapy for a variety of reasons
`(e.g., exhalation into the device, loading the device in incorrect
`orientations and inappropriate storage conditions may affect
`different DPls in different ways) (4, 15, 16).
`DPI formulations may either be drug mixed with a large
`particle size excipient (or diluent, e.g. lactose), to aid with powder
`flow, or it may consist of drug alone (3, 4). In all cases, the
`powder formulations are cleverly processed using proprietary
`techniques that are responsible for the DPI in question having
`apparently reproducible properties when tested in vitro under
`known, but constant, flow versus time profiles (e.g. for delivered
`dose and size distribution). From the development perspective,
`modern “passive” multidose DPIs (in which the patient provides
`the energy for powder dispersion) fall into two main categories.
`
`Figure 1. Photographs and diagrams of some of the inhalers described in the text. (a) Diagrammatic representation of MD! actuation into reservoir
`device prior to inhalation of the aerosol cloud from the reservoir. (b) Aradigm Corporation’s Smartmist "breath actuator” designed to house an
`existing MDl product and enable microprocessor control of its operation. Ensures actuation at “correct” point of inspiration, an appropriate
`inhalation rate and downloading to a PC for compliance checks (11). (c) Nektar’s Pulmonary Delivery System presently in Phase 11! trials with
`insulin as Pfizer-Aventis’ Exubera product. The device expels a puff of compressed air through a "transjector" inserted into a prepackaged blister
`containing spray dried insulin with proprietary excipients. The air puff creates an aerosol cloud for the patient to inhale through a mouthpiece
`from the integral reservoir atop the device. (d) Boehringer lngelheim’s Respimat "Soft mist inhaler.” A disposable, liquid self—metering device for
`small-volume (15 iii) aqueous and semi-aqueous solutions; lower part of diagram shows the conjunction of high pressure conduits leading to two
`opposed 10pm jets that create the fine mist as the spray streams converge; the company claim that the long duration (> 1 second) of the trigger-
`actuated aerosol cloud will enable "greater coordination and lung delivery” (24). (e) Diagram of Aradigm’s AERX operating principles; premetered
`drug solutions, in sterile blister-packs are pumped through an array of single—use, laser-drilled jets within the drug storage blister; results in efficient
`aerosolization and employs breath actuation and inspiration control, as well as patient education features. Device is expensive but has high delivery
`efficiency, humidity and temperature independence, and feedback on device usage during clinical trials (1 8, 19). (f) Picture of Chrysalis TechnologieS’
`prototype offering liquid metering, vaporization and condensation with claimed "pharmaceutical quality" for some drugs. Liquid is pumped and
`simultaneously subjected to microprocessor-controlled heating, during passage through a "capillary aerosol generator”; device can produce different
`aerosol droplet sizes commencing with MMADs in the submicron range (21, 22).
`
` WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 5 of 8
`
`WATSON LABORATORIES, INC. , IPR2017-01622, Ex. 1030, P. 5 of 8
`
`

`

`326
`
`PROCEEDINGS OF THE AMERICAN THORACIC SOCIETY VOL I
`
`2004
`
`the dependency of the fine particle dose on flowrate for marketed
`cromolyn sodium capsules in Spinhaler (UK product; lactose-
`free formulations) by using a constant throughput of 4 L of
`ambient room air at different flow rates. Whereas the delivered
`
`lations and are designed to minimize the effect that the patient
`has on the particle size distribution of the aerosolized drug dose.
`
`Novel Solution Metering Inhaler Devices
`
`Figures 1d, 1e, and 1f show some of the new solution metering
`devices that are currently in development for use with inhaled
`drugs. Several other companies are active in the area (Aerogen,
`Aradigm, Batelle Pharma, Chrysalis Technologies, Boehringer-
`Ingelheim, and Sheffield Pharmaceuticals; 26—34). All are devel—
`oping “propellant—free” inhalers that offer the advantage of liq—
`uid metering (for dosing precision) and an energy source to
`enable patient-independent, reproducible “active” aerosol pro-
`duction. Some devices are mechanical while others are electro-
`
`dose changed minimally (20 i 3 mg), the fine particle dose
`(dose < 5 pm aerodynamic diameter) differed by more than an
`order of magnitude in the range 30—100 L/minute (15, 16). Not
`surprisingly, all these dependencies are functions of both the
`inhaler and the formulation. For example, Turbohaler (With
`terbutaline sulfate) shows a strong dependence on flowrate,
`whereas Rotahaler and Diskhaler (albuterol sulfate) exhibit
`much less effect (14). To some extent this is predictable because
`the latter devices are rather inefficient and their designs ensure
`mechanical. The latter appear to have some added challenges
`only that they empty fairly well, not that they disperse powder
`from a development perspective, because of the need to demon-
`e