Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=iedd20

Expert Opinion on Drug Delivery

ISSN: 1742-5247 (Print) 1744-7593 (Online) Journal homepage: http://www.tandfonline.com/loi/iedd20

Considerations in establishing bioequivalence ofinhaled compounds

Irvin Mayers & Mohit Bhutani

To cite this article: Irvin Mayers & Mohit Bhutani (2018) Considerations in establishingbioequivalence of inhaled compounds, Expert Opinion on Drug Delivery, 15:2, 153-162, DOI:10.1080/17425247.2018.1381084

To link to this article: https://doi.org/10.1080/17425247.2018.1381084

Accepted author version posted online: 18Sep 2017.Published online: 21 Sep 2017.

Submit your article to this journal

Article views: 41

View related articles

View Crossmark data

REVIEW

Considerations in establishing bioequivalence of inhaled compoundsIrvin Mayers and Mohit Bhutani

Division of Pulmonary Medicine, Department of Medicine, University of Alberta, Edmonton, AB, Canada

ABSTRACTIntroduction: Generic inhalers are often perceived as inferior to their branded counterparts; however,they are safe and effective if they can meet the regulatory requirements. The approach to assessbioequivalence (BE) in oral dosage form products is not sufficient to address the complexities ofinhalational products (e.g., patient-device interface); hence, more considerations are needed andcaution should be applied in determining BE of inhaled compounds.Areas covered: This review outlines the evaluation process for generic inhalers, explores the regulatoryapproaches in BE assessment, and highlights the considerations and challenges in the current in vitroand in vivo approaches (lung deposition, pharmacokinetic, pharmacodynamic/clinical studies, andpatient-device interface) for establishing BE of inhaled compounds.Expert opinion: The ultimate goals in this field are to establish uniformity in the regulatory approachesto speed the drug submission process in different regions, clear physicians’ misconception of genericinhalers, and have meaningful clinical endpoints such as improvement in patient quality of life whencompared to placebo and brand name drugs. As inhalational drugs become more common for otherindications such as antibiotics, the technologies developed for inhaled compounds in the treatment ofchronic pulmonary diseases may be extrapolated to these other agents.

ARTICLE HISTORYReceived 13 April 2017Accepted 14 September 2017

KEYWORDSBioequivalence; inhaledcompounds; genericinhalers; chronic obstructivepulmonary diseases; asthma

1. Introduction

Generic drugs are often perceived as inferior to their brandedcounterparts; however, generic and brand name drugs haveidentical active ingredients. In Canada, generic drugs mustdemonstrate equivalence by meeting Health Canada’s stan-dards for bioequivalence (BE) [1]. Bioequivalent drug formula-tions have the same bioavailability; therefore, new clinicalstudies are not required. All generic manufacturers mustmeet standards for good manufacturing practices; theseinclude quality standards for ingredients, assays, manufactur-ing processes, and facilities. Moreover, Health Canada setsmore stringent criteria for highly toxic drugs or those withnarrow therapeutic range (e.g. antirejection drugs).

Unlike innovative new drugs, marketing authorization for ageneric product can be obtained by an abbreviated new drugsubmission. There are regulatory approaches to ensure thatthe generic product is pharmaceutically equivalent and bioe-quivalent to the Canadian reference product (CRP), which isusually the branded product marketed by the innovator of thedrug, using a series of comparative studies (e.g. bioavailability,pharmacodynamics [PD], or clinical studies) [1].

The approach to assess BE in oral dosage-form products(e.g. area under the curve [AUC] [2]) is not sufficient to addressthe complexities of inhalational products because AUC forinhaled products does not accurately reflect the product’sbiological activity. Moreover, systemic assessments that areused for oral dosage-form products may not be applicable tocompounds with localized targets such as inhalational

products that are used to treat bronchoconstriction and/orlocalized inflammation in the airways [3,4]. As such, the useof a specific target end point (e.g. eosinophils or nitric oxide) issuggested over the use of secondary target measurementssuch as forced expiratory volume in 1 s (FEV1) when evaluatinganti-inflammatory inhaled medications. In addition, thepatient-device interface should be assessed for inhalationalproducts, which adds complexity to the evaluation process.

A Canadian draft guidance for BE considerations for inhaledcorticosteroids (ICS) suggested that the primary end point in aproposed study should be a marker of eosinophilic airwayinflammation [5]. While this Canadian guidance document isintended for ICS, it can be used to frame the general guidancefor BE assessment of inhaled compounds in chronic respiratorydiseases such as chronic obstructive pulmonary diseases(COPD), which is currently lacking.

The purpose of this review is (i) to clear the misconceptionthat generic inhalational products are inferior to their brandedcounterparts by describing the process of BE evaluation, and(ii) to emphasize the challenges in establishing BE of inhaledcompounds by reviewing the in vitro and in vivo approachesand highlighting the considerations in each approach.

2. Generic inhalational products - what is BE?

According to Health Canada, pharmaceutically equivalent isdefined as, when compared with another drug, a new drugcontaining identical amounts of the identical medicinal

CONTACT Irvin Mayers imayers@ualberta.ca Division of Pulmonary Medicine, Faculty of Medicine, University of Alberta, 3-126 Clinical Sciences Building,8440 112 Street NW, Edmonton, AB T6G 2B7, Canada

EXPERT OPINION ON DRUG DELIVERY, 2018VOL. 15, NO. 2, 153–162https://doi.org/10.1080/17425247.2018.1381084

© 2017 Informa UK Limited, trading as Taylor & Francis Group

ingredients in comparable dosage forms; however, the newdrug does not need to contain the same non-medicinal ingre-dients [6]. In Canada, generic drugs must meet Health Canada’sstandards for BE, in which bioequivalent drug formulationsmust have the same bioavailability [1]. Bioavailability can bedetermined using the AUC of the drug’s blood concentrationvs. time plot. For generics to be considered bioequivalent, theAUC and the maximum observed concentration (Cmax) must beno less than 80% or no more than 125% of the referenceproduct [1]. Similarly, the 90% confidence interval (CI) of theAUC and Cmax must also fall within 80–125%; for the CI to fallwithin this range, variance is generally <5% (Figure 1). Thus,although it initially appears that the comparison between gen-eric and reference products is broad (i.e. AUC falls within80–125%), requiring that the 90% CI also falls within thisrange ensures that any differences between the products is infact very narrow.

The excipients in the inhalational products may differbetween the reference and generic, which is an importantconsideration for patients with allergy or sensitivity to excipi-ent(s) (e.g. lactose, gluten, sulfites, or tartrazine). Differences inexcipients may become more important when consideringdrugs with pediatric indications. Generic and reference pro-ducts may also differ in color, shape, or markings. It should benoted that interchangeability is not the same as BE. In Canada,BE is based on comparative bioavailability; however, eachprovince or territory determines interchangeability based on

its own policies or regulations [1]. Therefore, after a genericdrug is deemed to be bioequivalent with the innovator, aprovincial regulator may also deem that this generic drugcan be substituted for the innovator product without patientor prescriber consent.

When determining BE of oral or inhaled drugs, AUC andCmax reflect multiple factors, such as release of the activeingredient from the formulation, and the rate and extent ofabsorption from the gastrointestinal (GI) tract. BE studies are asurrogate marker for clinical effectiveness and safety data, andit can also be demonstrated through other comparative test-ing such as PD studies. For oral drugs, if the blood concentra-tions are the same, then safety and effectiveness are predictedto be the same; however, inhalational products require othercomparative analyses.

Health Canada, US Food and Drug Administration (FDA),and European Medicines Agency (EMA) all have different eva-luation processes for establishing BE of inhalational products[7]. Overall, Health Canada and FDA endorse in vitro studies asthe starting approach, while EMA accepts an in vitro-onlyapproach. All three agencies aim to make safe, effective, andless costly medications available to their respective popula-tions, but it is concerning that there is a great degree ofdiversity in their approaches to the same products.

2.1. Health Canada

The Canadian Food and Drug Regulations has strict guidancethat address physical equality (e.g. particle size), which isnecessary when considering inhaled products because physi-cal properties of the medicinal ingredients could potentiallycause differences in the safety and efficacy profiles [6,8].

Presently, Health Canada does not have guidance for long-acting generic inhaled compounds in the treatment of COPD.The guidance to determine equivalence of subsequent entryshort-acting beta agonist (SABA) stated that reliance solely onin vitro data is not suitable due to unestablished correlationsbetween in vitro and in vivo data [9]. Further to this guidance,the Scientific Advisory Committee on Respiratory and AllergyTherapies (SAC-RAT) strongly recommended a 1-year safetystudy to support registration of a subsequent entry SABA[10]. In its guidance document relating to the pharmaceuticalquality of inhalation and nasal products, SAC-RAT recom-mended that physical properties, such as particle size andextractables and leachables, be tested for these products [11].

Article highlights

● Generic inhalers are not inferior to branded inhalational products interms of effectiveness, safety, and quality; they are safe and effectiveif they can meet the regulatory requirements.

● Regulatory approaches for the BE assessment of oral dosage formproducts are not sufficient for inhalational products, and cautionshould be applied in determining BE of inhaled compounds.

● In the evaluation approaches, aerosol behaviour should be studied,and lung deposition studies are suggested in the current approach toassess BE of inhaled compounds.

● Clinical trials should be of adequate duration (e.g., long-term studies),which include evaluations of both safety and effectiveness for eachcompound.

● There are still challenges in establishing BE of inhaled compounds(e.g., no uniform regulatory approaches).

This box summarizes key points contained in the article.

Figure 1. For bioequivalent drugs to be accepted, 90% confidence interval of the area under the curve must fall within 80%–125% of the branded drug.

154 I. MAYERS AND M. BHUTANI

In 2011, the draft guidance for treatment of asthma usingICS delivered via metered dose inhalers (MDIs) and dry pow-der inhalers (DPIs) recommended the measurement of primaryendpoints that reflect the action and effect of ICS (i.e. inflam-matory outcome) [5]. Both generic and CRP should have 50%reduction in eosinophils when compared with placebo. Theguidance strongly recommends the incorporation of second-ary end points, such as FEV1, and an asthma control question-naire. Generic and CRP should have a mean FEV1 that isgreater than placebo by at least 10%. The 90% CI of therelative means must be within 80–125% in order to demon-strate equivalence.

Current evidence in COPD research indicates that changesin drug-device combinations, different dose and formulations,along with specificities of study design, can have a significanteffect on measurable outcomes and associated BE [12]. Theasthma-ICS draft guidance indicates that safety considerationsfor subsequent entry ICS products should be addressed viaappropriate pharmacokinetic (PK) study and comparable invitro characteristics [5]. In vitro characteristics include similaractive and inactive ingredients (i.e. propellant mix), and similarperformance characteristics such as particle size distributionand velocity of the aerosol plume. PK studies should be asingle-dose study at the upper limit of the dosing range, andAUC, Cmax, observed time at which Cmax is reached (tmax), half-life, and terminal elimination rate constant should be deter-mined. PD study should be conducted in healthy adult volun-teers over asthmatic patients in order to preclude thepotentially confounding effects of past or current steroid ther-apy, and variability in the degree of airway inflammation andobstructive impairment, which could act to reduce the powerof the analysis. A study duration of at least 3 weeks is neededif parallel study design is used because time is needed toreach the therapeutic plateau. The dosage of CRP should bebelow the therapeutic plateau but still on the steep slope ofthe dose–response curve. Alternatively, a balanced crossoverschedule of treatments with washouts between each dose isrecommended.

The asthma-ICS guidance and the latest scientific literatureprovide a scientific framework for a protocol to exhibit BE ofsubsequent entry long-acting compounds in the treatment ofCOPD. In 2009, SAC-RAT released a general, preliminary gui-dance for subsequent entry long-acting beta agonists (LABAs),long-acting muscarinic antagonists (LAMAs), and fixed-dosedrug combinations [13]. This guidance is preliminary andrequires further consultation, but it should be considered asa starting point for future guidance documents for LABAs.

The preliminary guidance stated that, for subsequent entryLABAs, the basic data requirements include PD/clinical study,PK study (for safety), and in vitro characterization of the parti-cles and the delivery device [11]. For PD/clinical study, a mini-mum of 1 year is recommended in efficacy trial, and a 1-yearsafety study is not necessary for LABAs indicated for COPDunless the generic was submitted as a new drug submission.SAC-RAT recommends including patients with stable COPDrequiring no intervention in the preceding 6 weeks andStage 2–3 severity per the Global Initiative for ChronicObstructive Lung Disease, as well as healthy volunteers/

controls. A randomized three-arm crossover design (i.e. test[generic], reference, placebo) with a washout period of 72 hand an adequate run-in period is recommended. The PD studydesign should be the same as that outlined in the asthma-ICSdraft guidance, with AUC of the FEV1 as the primary efficacyend point. To establish therapeutic equivalence (TE), the 90%CI of test/reference ratio of mean change in the primaryefficacy end point should be within 80–125%. SAC-RAT alsorecommends including FEV1 data for claiming alteration ofdisease progression, COPD-specific quality of life (QoL) ques-tionnaires for claiming reduction of COPD symptoms, 6-minwalk test results for claiming improvement of exercise capa-city, and well-defined moderate-to-severe COPD exacerbationsfor claiming reduction of COPD exacerbations. For PK study,QTc interval, heart rate, serum potassium, and glucose levelsshould be tested as additional safety parameters.

2.2. US Food and Drug Administration

At present, the FDA adopts a stepwise approach to establishBE of generic inhalers. It uses the weight of evidenceapproach, meaning that each drug application should incor-porate qualitative and quantitative formulation similaritieswithin acceptable margins, data from PK studies for the assess-ment of equivalent systemic exposure, and data from studiesfor the evaluation of equivalent local delivery [14,15].

The draft guidance released by the FDA in 2003 for thedesign of bioavailability and BE studies for nasal aerosols andnasal sprays for local action gave key considerations to drugparticle size and distribution patterns, which are dependenton the drug substance, formulation, and device characteristics;similar biophysical considerations would be applicable toinhaled compounds [16]. In recent years, instead of releasinga general guidance document for determining BE betweentest and reference for all LABAs and LAMAs, the FDA releaseddraft guidance that are tailored for individual inhaled com-pounds and their combinations; draft guidance are availablefor fluticasone/salmeterol [17], indacaterol maleate [18], ipra-tropium bromide [19], aclidinium bromide [20], beclometha-sone dipropionate [21], ciclesonide [22], formoterol fumarate/mometasone furoate [23], budesonide/formoterol fumaratedihydrate [24], albuterol sulfate [25], and levalbuterol tartrate[26]. All of these draft guidance recommend a combined invitro and in vivo approach for establishing BE of the test andreference inhalers. Generally, in vitro analysis should include ameasurement for single actuation content and aerodynamicparticle size distribution, with specific requirements for thedevice design and formulation. A PK study with crossoverdesign in healthy volunteers is recommended, and equiva-lence is based on AUC and Cmax for the inhaled compound.PD/clinical study should have a parallel or crossover designwith washout periods, and it should be conducted in COPD orasthmatic patients. Overall, FEV1 is recommended as the pri-mary end point, but there are specific endpoints for albuterolsulfate, and levalbuterol tartrate.

In vivo tests in humans measuring concentration-time pro-files are the most thorough approach in terms of accuracy,sensitivity, and reproducibility for determining BE. However,

EXPERT OPINION ON DRUG DELIVERY 155

for drugs that are not intended to be absorbed into thesystemic circulation, such as inhaled compounds, BE shouldbe assessed by measurements intended to reflect bioavailabil-ity at the site of action. The FDA recognizes that the assess-ment of BE for locally acting products has presented a uniquescientific challenge, and that the means by which certain testsare carried out can have a significant effect on outcomes. PKstudies cannot easily address the extent of pulmonary deposi-tion and the ratio of central to peripheral lung deposition. Assuch, lung deposition study is suggested to accompany PKresults. In addition, trials of appropriate statistical power forpreselected endpoints should be conducted to establish TEand determine the clinical effect of the compound.

2.3. European medicines agency

The EMA has considered the underlying scientific challengesassociated with inhalational products. Current guidelines out-line that BE between two inhaled products cannot be deter-mined, but appropriate evaluation may lead to comparable TE.The guidance document by the EMA considers a stepwise invitro and in vivo approach for determining equivalence.Approval of a generic inhaler based solely on in vitro testingis possible if the following criteria are met: identical dosageform, aerosol particle behavior, identical delivered dose(±15%), inhaled volume through the device to be ±15% ofreference device, same resistance of the inhalation device toairflow (±15% of reference device), and similar handling of theinhalation devices to release the required amount of activeingredient [27].

If a generic inhaler fails in vitro assessment, TE can beestablished through in vivo pulmonary deposition using ima-ging studies or PK evaluation, with technical considerationsacknowledged by the EMA, and PK similarity (i.e. upper limit of90% CIs for the test/reference ratio is 125%) is desirable [7]. Iflung deposition is not sufficient, clinical equivalence can beestablished through demonstrating both equivalent efficacyand safety. Alternative approaches to PD safety analysis can beconsidered, such as assessing the biochemistry, hematology,and electrocardiogram (ECG) activity at maximal effect (fastingstate) [7]. Safety effects are generally not seen with approvedand lower doses of SABAs and LABAs; thus, doses greater thanthe maximal approved dose may need to be assessed toobserve known class effects. Vital signs, ECG, serum potassium,and blood glucose should also be evaluated. Specific toLAMAs, local adverse events (e.g. dry mouth) should beconsidered.

2.4. Summary of regulatory approaches

The brief review of regulatory approaches described above issummarized in Table 1, and it highlights the complexity ofdetermining BE or TE between test and reference productsand the inconsistency between different regulatory agenciesregarding interpretation of similar scientific challenges. Drug-device design and patient-device interfaces may add morecomplexity to these evaluation approaches. In order to furtherestablish BE standards for locally acting inhaled products,more work is needed to develop well-defined protocols with

established end points for comparative clinical studies,develop and validate in vitro–in vivo correlations, and con-struct statistically justified criteria for declaring BE [28].

3. How to establish equivalence – in vitro studies

In vitro studies can be used to predict the in vivo behavior ofinhalational products such as lung deposition. Typically, invitro studies characterize the physical properties of the inhaledcompound, which include, for example, particle size (medianand distribution), solubility, shape, density, rugosity, charge,crystallinity, and velocity of the aerosol plume [11].Particularly, the FDA recommends that in vitro analysisincludes a measurement for single actuation content and foraerodynamic particle size distribution [17–26]. The propertiesto be tested vary and depend on the mode of inhalation.

Establishing BE for generics purely on the basis of in vitroresults is insufficient because there are different mechanismsinvolved in in vitro sampling and in vivo deposition for

Table 1. Requirements for bioequivalence from Health Canada, FDA, and EMA.

Agency Requirements

HealthCanada

Must meet all of the following:● In vitro similarity

○ Similar active ingredients○ Similar performance characteristics (e.g. particle size distri-bution, velocity of aerosol plume)

● Systemic PK similarity○ Single-dose study at the upper limit of the dosing range○ Similar AUC, Cmax, tmax, half-life, and terminal eliminationrate constant

○ Lung deposition study● PD or clinical similarity

○ Mean FEV1 greater than placebo by at least 10%; 90% CI oftest/reference ratio of the mean change should be within80–125%

○ Inflammatory outcome – 50% reduction in eosinophilscompared to placebo

○ Efficacy trial of at least 1 year, with AUC of FEV1 as theprimary end point

○ 1-year safety study○ COPD-specific QoL questionnaires○ 6-minute walk test○ Reduction of COPD exacerbations

FDA

EMA Must meet one of the following:● In vitro similarity

○ Identical dosage form○ Identical aerosol particle behavior○ Identical delivered dose (±15%)○ Inhaled volume through the device (to enable enoughactive ingredient into the lungs) to be ±15% of reference

○ Same resistance of the inhalation device to airflow (±15% ofreference)

○ Similar handling of the inhalation devices between test andreference for releasing the required amount of activeingredient

● Systemic PK similarity○ Similar AUC, Cmax

○ Upper limit of 90% CIs for test/reference ratio is 125%○ Lung deposition study

● PD or clinical similarity○ Equivalent efficacy and safety○ Alternatives to PD safety analysis: biochemistry, hematology,electrocardiogram activity at maximal effect

AUC: area under the curve; CI: confidence interval; Cmax: maximum observedconcentration; COPD: chronic obstructive pulmonary disease; EMA: EuropeanMedicines Agency; FDA: Food and Drug Administration; FEV1: forced expira-tory volume in 1 s; PD: pharmacodynamic; PK: pharmacokinetic; tmax: observedtime at which Cmax is reached.

156 I. MAYERS AND M. BHUTANI

inhalational products. First, in vitro sampling involves principlesof impaction based on sequential stages, each of which hasincreasing linear velocity at a constant volumetric flow rate. Inthe principles of inertial impaction, large aerosol particles aredeposited on the walls of airway conduit and the impactiontends to occur where the airway direction changes; in contrast,small particles have less inertia and are more likely to be carriedaround corners and continue in the path of the airflow. In theairways of the lungs, linear velocity largely decreases towardthe periphery due to increasing cross-sectional area at anygiven flow rate, and variable cyclic airflow velocities areinvolved, which give rise to a different mechanism for in vivodeposition compared to in vitro sampling [28] (Figure 2).

Second, the in vitro predictive method would at best modelthe physical properties of the airways because impactors andtheir inlets have fixed dimensions with rectangular air ducts,whereas airways of the lungs are flexible with variable dia-meters and lengths, and possibly changing angles of bifurca-tion upon inhalation [29]. Despite the above concerns, inertialimpaction has been used as the standard method for deter-mining the aerodynamic particle size distribution of pharma-ceutical aerosols for approximately 50 years [30].

Finally, it is difficult for in vitro models to address all vari-ables involved in aerosol generation. In vitro models are gen-erally static and will not allow for considerations such aspatient-device interface and the inherent variability associatedwith the disease, which is accounted for in robust clinical trialprograms of the reference product.

Given these differences in sampling and deposition, in vivostudies and clinical tests are needed for establishing BE ofinhaled compounds.

4. How to establish equivalence – in vivo studies

4.1. Lung deposition studies

Lung deposition studies are often used as a means to under-stand the in vivo performance of generic inhalers/formulationsin relation to the established products. These studies are

suggested to complement PK assessments because theyallow quantification of whole and regional lung depositionas well as extra-thoracic deposition [31]. PK studies, whichevaluate systemic exposure, and conventional pulmonaryfunction tests such as FEV1 do not provide information aboutregional drug deposition [32]. Moreover, data from lungdeposition studies positively correlate with clinical efficacy ofinhaled drugs [33].

Efficient drug targeting to airway regions is influenced byfactors such as aerosolized particle size distribution and den-sity, efficiency of aerosolization technology, timing of theaerosolization during inspiration, inhalation flow rate-timeprofile, inhalation volume, and geometrical configuration ofairways [32,34,35]. Generally, particles with size <5 μm can bedistributed deep into the lung with good clinical response[35]. One study reported that there is a high degree of varia-bility in respiratory tract deposition of 3-μm particles inpatients with different respiratory symptoms due to differ-ences in flow rates [36]. Particles with size <2.5 μm are mainlydeposited into the capillary-rich alveoli where they may exertno PD effect and are rapidly absorbed [33,35]. The finestparticles with size <1 μm are mostly exhaled if they are notaggregated and/or not migrating to the lung walls [35].

Lung deposition studies commonly use radionuclide orradiolabel imaging, which allows for noninvasive quantifica-tion of whole and regional lung deposition. Care should beexercised with the use of radionuclide labeling in lung deposi-tion studies, which includes correction for differences inbreathing patterns and tissue attenuation, development ofradiolabelling methods and lab-to-lab variation, validationtesting for demonstrating that the addition of radiolabel hasnot changed the chemical and physical properties of the drug,and changes in particle size distributions over the shelf life ofthe product, which may affect deposition [37].

Most lung deposition studies to date have used 2D (e.g.gamma scintigraphy) and 3D imaging methods, such as sin-gle-photon emission computed tomography (SPECT) and posi-tron emission tomography (PET) [35,38,39]. Briefly, gammascintigraphy uses the radionuclide 99mTechnetium (99mTc)

Figure 2. A schematic showing, in the airways of the lungs, linear velocity decreasing towards the periphery due to increasing cross-sectional area at any given flow rate.

EXPERT OPINION ON DRUG DELIVERY 157

and it is the most commonly used method in lung depositionanalysis [35,40]. SPECT allows full 3D reconstruction of thelungs and it generally uses 99mTc as the radiolabel [40]. PETuses a positron-emitting radionuclide such as 11C, which canbe incorporated in the molecule’s structure without changingits chemical properties, allowing for highly accurate measure-ments of deposition and clearance; however, this method islimited by short half-lives of the positron emitters and cost[35,40,41]. There are currently devices with a combined PET/CTfeature which allows for accurate incorporation of PET imageswith anatomical data [35].

Recently, magnetic resonance imaging (MRI) has been con-sidered in lung deposition analysis. MRI uses nuclear magneticresonance and thus, no radiolabel is required. However, con-trast agents are usually added to aerosol formulation formeasurements, which limits most of these studies in animals[35]. There are also novel simulation methods such as func-tional respiratory imaging (FRI), which is based on computer-ized tomography analysis and computational fluid dynamics.FRI simulates outcomes such as ventilation, lung deposition,and perfusion of airway blood vessels in a patient-specificmanner. The enhanced sensitivity with FRI allows it to detectclinically relevant changes on a regional level and provideregional information in terms of deep lung deposition (distaland peripheral airways) and bronchodilation [32].

Overall, 2D scintigraphy provides a simple index of regionallung deposition with limited information on the anatomicalsite of deposition because 2D images of the lung representthe compression of a complex 3D architecture, consisting of amixture of large and small conducting airways and alveoli, intotwo dimensions [31]. 3D imaging methods provide more infor-mation on regional lung deposition than 2D imaging (e.g.distribution in different planes) but they are more technicallychallenging than 2D scintigraphy.

An example of lung deposition study is the evaluation ofRespimat® Soft MistTM Inhaler (a propellant-free inhaler) andthe Handihaler® for inhaled tiotropium. Respimat® Soft MistTM

Inhaler is an alternative to the Handihaler® for the delivery ofinhaled tiotropium (a long-acting anticholinergic bronchodilator).Scintigraphic studies showed that there was greater lung deposi-tion with the Respimat® platform when compared with a pres-surized MDI (pMDI); and the mechanism of drug delivery of theHandihaler® suggests a lesser degree of lung deposition com-pared to the Respimat® platform [42]. Using the EMA’s equiva-lence algorithm, equivalence was not established between theRespimat® and Handihaler® devices based on deposition data.

When assessing BE using lung deposition studies, the diffi-culties of establishing dose–response relationships for inhaledasthma and COPD drugs should be considered. Clinicalresponse to a bronchodilator reaches a plateau (ICS haverelatively flat dose–response curves), and further drug deposi-tion in the airways does not further augment the response[43]. Deposition studies are generally conducted using optimalinhalation techniques in order to maximize lung deposition,whereas inhalation techniques are not as well controlled inclinical studies (consider breathing patterns) [44]. Also, deposi-tion studies have often been carried out in healthy volunteers,in whom aerosol deposition and its variability differ frompatients with lung disease.

Given the benefits of lung deposition studies, they shouldbe considered an important factor in assessing and establish-ing BE of inhalation compounds.

4.2. PK studies

Two inhaled products would be expected to show equivalenceif the same dose is deposited in the lung with delivery to thesame central and peripheral parts of the lung, and the drugenters the same pulmonary cells at the same rate with thesame receptor kinetics.

PK analyses utilize measurements from serum (plasma) orurine [40]. For inhaled drugs, the sampling site for PK studies isa compartment that is downstream of the site of action (lung);thus, results from PK studies may not adequately reflect thetherapeutic effect resulting from drug–receptor interactionwithin the lung. PK studies can estimate differences in totalsystemic delivery of inhaled products via AUC; however, twoformulations could lead to a similar AUC calculation over adosing interval even if the central and peripheral deposition isdifferent [40]. Relative changes in Cmax or partial AUCs mayprovide a measure of different rates of absorption from differ-ent areas of the lung.

Previously, PK studies demonstrated that hydrofluoroalkaneformulations of beclomethasone dipropionate had greaterpulmonary deposition than the chlorofluorocarbon product,which agreed with the superior efficacy shown in PD studies[45]. In another study, two DPI-based combinations of flutica-sone propionate and salmeterol showed significant differencesin PK profiles, but clinical studies evaluating the pulmonaryeffects did not suggest any differences [46,47]. In addition, thestudy that compared Respimat® and Handihaler® for thedelivery of tiotropium showed that BE was not establishedbecause the 90% CI was below the pre-specified acceptanceinterval of 80–125% [48]. Despite the lower systemic exposureof tiotropium with Respimat®, both devices demonstratedsimilar bronchodilator efficacy in clinical studies.

PK studies need to be designed carefully to account forregional absorption and potential differences in absorptionrates. Systemic delivery via GI and pulmonary routes can beseparated by blocking GI absorption with the administrationof oral charcoal during lag time of the absorption phase [40].While not sufficient on its own, an appropriately designedcomparative PK study provides great insight into the clinicalsafety of a generic inhaler. Plasma levels are usually moreclosely related to side effects or adverse events of inhaleddrugs, which are mainly systemic effects occurring down-stream of the central compartment; thus, these effects canbe evaluated through comparative PK studies. In addition toabsorption, attention should also be given to excretion of thedrug as it can affect systemic absorption (e.g. renal clearance).

Pulmonary residence time is another consideration in PKstudies that is applicable to some formulations. Pulmonaryresidence time affects the degree of pulmonary selectivityand it should be similar amongst bioequivalent formulationsbecause thin alveolar membranes and significant pulmonaryblood flow allow solubilized drugs to be quickly absorbed intothe systemic circulation. Depending on the chemical proper-ties of the inhaled compound, Cmax and possibly tmax may

158 I. MAYERS AND M. BHUTANI

decrease with decreasing absorption rate; therefore, PK stu-dies may provide insights into potential differences in pul-monary residence time [49]. In addition, these studies shouldbe carried out during PK steady state to reduce the effects ofinitial plasma fluctuations.

4.3. PD studies/clinical trials

PD studies report on the clinical efficacy and safety of genericinhalational products. Large phase III clinical programs maynot be needed to establish safety if lung deposition and PKstudies have stringent evaluation criteria and if analysis showscomparable results to the reference product. However, theFDA encourages clinical studies in patients for inhalationaldrugs due to their complexity and associated delivery systems;discrepancies have been observed amongst in vitro studies, PKstudies, and safety [50].

While healthy volunteers are usually deemed suitable forPK studies, in studies of inhaled drugs conducted in patients,pulmonary disease significantly influences the deposition ofthe drug due to factors such as narrowing of conductingairways, mucus, and plugs. This issue can be addressed byadopting a crossover design, which allows for increasing studypower and reducing the number of patients needed in thestudy [51,52].

In determining the clinical efficacy of inhaled products, twokey effects should be evaluated – bronchodilation andbronchoprotection. Bronchodilation may be assessed throughFEV1 measurement before and after delivery of inhaled drug atsteady-state conditions [51,52]. Time-dependent bronchodila-tion may be evaluated through repeated FEV1 measurementsand generation of the area under the FEV1-time curve.Bronchoprotection may be evaluated through bronchial chal-lenge studies, in which a bronchoconstrictive agent is intro-duced at increasing dosage, and FEV1 is repeatedly measureduntil a 20% decrease is observed compared to baseline, thusdefining a provocative concentration; then, comparison of thisvalue before and after delivery of inhaled drugs allows forcomparison of their therapeutic effects [53].

While most compounds used in the treatment of COPDmay exhibit bronchodilation and bronchoprotection effects,reference products are typically studied for other outcomessuch as degree of daily symptoms, rate of exacerbations, needfor hospitalization, and health-related QoL. Safety evaluation isimportant in establishing BE of inhaled compounds. For exam-ple, two different formulations of tiotropium (Respimat® com-pared with Handihaler®) were compared in a 3-year study interms of time to all-cause mortality and time to first COPDexacerbation [54]. Results demonstrated that tiotropium deliv-ered by Respimat® was non-inferior to tiotropium delivered byHandihaler® with respect to the risk of all-cause mortality andnot superior to Handihaler® with respect to the risk of firstexacerbation.

For considerations in PD studies, caution must be appliedwhen extrapolating the effects of a generic inhaler to all out-comes associated with the reference product. Trials of appro-priate length and sample size should be conducted in order toprovide sufficient statistical power for accurate comparisonsbetween a generic and reference inhalational product.

Moreover, careful consideration should be given to dosagein the design of trials. It is desirable to test drugs on thesteepest part of their dose-response curve because it is usually‘S’-shaped; thus, evaluating BE on either plateau of the curvewould be inaccurate. Evaluating two dose strengths of the testand reference drugs may improve accuracy. For statisticalmethods in evaluating BE of inhaled compounds (i.e. compar-ing dose–response curves), the Finney parallel regression ana-lysis and the nonlinear Emax model can be used [55,56].

Given the complexities of determining equivalencebetween a test and a reference product, clinical studies arenecessary to establish comparable efficacy and safety, espe-cially if PK analysis is not feasible.

4.4. Patient-device interface

Treatment in the pulmonary area is heavily device-dependentand the device component of the drug-device combinationadds complexity to BE assessments. There are two essentialfactors in achieving reproducibility – design of the device andthe patient’s technique to correctly use the device.

4.4.1. Design of inhaler deviceInhaler devices have been designed based on patient physiol-ogy and the molecule’s physical characteristics. In terms ofpatient physiology, efficacy of the inhaled drug can beaffected by the type of disease even if the same device wasused. For example, in the study of Sharma et al., the Cmax oftiotropium administered by the Respimat® platform is 50%lower in patients with asthma compared to COPD [57]. Forphysical characteristics, the device and the drug are not easilyseparable in suspensions or in dry powder devices; thus, anextensive series of physical property measurements arerequired before any BE testing can be considered.

In the design of a device, patient needs are importantconsiderations and these can be placed into three categories– use, interface, and perception. In terms of patient use, thefundamental functions of the device (i.e. dose metering, flowresistance, handling, and sequence) should be considered.Differences in airway flow rate between healthy and diseasestates should be accommodated in the design. It should benoted that suboptimal flow in patients with COPD does notallow complete inhalation from DPI, leading to poor clinicaloutcome [58]. For the patient–device interface, the function-alities (e.g. dose counter, graphical instructions for ease of use,intuitive operations, and locking mechanisms) should be con-sidered. Patient perception may include patient preferencemeasures but they are difficult to quantitate. While risk man-agement analysis can be used to determine the requiredstudies to address these concerns, little consensus has beenreached in a regulatory framework to address them.

4.4.2. Device handling by patientsIncorrect inhaler usage and non-adherence to therapy havebeen recognized as major factors in uncontrolled asthma orexacerbation of COPD [59]. An observational study found thatup to 76% of patients were reported to improperly use deliv-ery devices [60]. Another study showed that upon switchingdevices, the reported rate of successful treatment decreased

EXPERT OPINION ON DRUG DELIVERY 159

from 34.3 to 19.7% [61]. In particular, when switching frombranded to generic inhalers, if patients with asthma or COPDdo not receive appropriate training to use the device, this canresult in poor clinical outcomes, increased healthcare costs,and poor patient–physician/patient–pharmacist relationship[59,62]. It has been demonstrated that devices with lowervelocity of the aerosol could allow for greater lung depositionwhen compared with conventional pMDIs, possibly due tobetter coordination for patients (i.e. better technique) [42]. Ingeneral, factors such as patient age, severity of airway obstruc-tion, and degree of training have been associated with poorpatient-device interfaces [63]. For example, a 53% increase inlung deposition was observed in patients who were trained onthe Respimat® device [64].

Given these complexities, a true generic inhalational pro-duct cannot be evaluated independent of the device [65]. Infact, the patient-device interface is increasingly being consid-ered in the design of clinical trial programs.

5. Conclusion

This review outlined the evaluation process for generic inhala-tional products and the current in vitro and in vivo approachesfor establishing BE of inhaled compounds, and highlighted theconsiderations and challenges in these approaches. There arestill multiple considerations and challenges in establishing BEbetween generic inhalers and their corresponding referencedrug products. At present, PK studies are commonly used todetermine BE for systemically acting drug products; however,such an approach is unlikely to be sufficient to establish BE ofinhaled drugs because their intended actions and deliveriesare at the local sites and consequently, they do not rely onsystemic circulation. As such, lung deposition studies arerecommended to accompany PK analyses. For inhaled medi-cations, drug-device combination and device-patient interfaceare both critical considerations in the development of theseproducts. The same active ingredient can be formulated insimilar or different inhalational devices but their aerodynamicperformance may vary, which may lead to dissimilar drugdelivery to the airways. Thus, it is imperative that comparativeclinical efficacy and safety be properly evaluated with a robustclinical trial program to ensure that patients achieve thedesired clinical outcome with the generic inhaler.Furthermore, safety evaluation should be established vialong-term clinical studies in patients with the disease state.

6. Expert opinion

Currently, Health Canada, the FDA, and the EMA have differentsets of approaches for BE assessment. There should be aunified, general set of evaluation approaches as a startingpoint for establishing BE for all inhaled products. Once thesebasic requirements were met, it is recommended that the testcompound be assessed through drug-specific and/or device-specific approaches. Ideally, a comprehensive documentshould be developed through a collaboration of several reg-ulatory agencies to cover all important issues, which wouldensure that future products meet the high standards and be

adopted internationally and allow for a streamlined process toexpedite market entry for generic inhalers.

In addition to formulation, device characteristics andpatient-device interface are important factors that affect clin-ical response to orally inhaled drug products. However, clini-cally important differences in patient–device interactions havenot been established; thus, clinical studies should consideradding this end point to the study design.

In the field of generic inhalational products, the ultimategoals are to establish uniformity in the evaluation approachesto speed the drug submission process in different regions,clear physicians’ misconception of generic inhalers, and havemeaningful clinical endpoints such as improvement in patientQoL when compared to placebo and brand name drugs.

In the coming years, there will be increased pressure in theapproval of generic inhalers to gain market entry, and regulatoryprocess should not be the limiting reason. In the approach toevaluate BE, lung deposition studies will become the center ofattention. The methodology to assess lung deposition of sepa-rate compounds in a combination inhaler will be of interestbecause currently, there are no such techniques available. Itcan be foreseen that lung deposition studies will be increasinglyused for BE assessment, and with the advancement in medicalimaging technology and computational capabilities, lungdeposition analysis may be sufficient to establish BE once invitro studies are completed. In addition, there will be combina-tion inhalers with more than two compounds, which makesdrug-device design and BE study designs more challenging.

As inhalational drugs become more common for otherindications (e.g. antibiotics, mucolytics), the technologiesdeveloped for LABAs and ICS may be extrapolated to theseother agents.

Funding

This paper was not funded.

Declaration of interest

I Mayers has received an honorarium from Boeringher Ingelheim for CMErelated to generic inhalers. Medical writing assistance was used in thedevelopment of this paper and was carried out by Jane Cheung (SAGEMedica Inc.). The authors have no relevant affiliations or financial involve-ment with any organization or entity with a financial interest in or finan-cial conflict with the subject matter or materials discussed in themanuscript. This includes employment, consultancies, honoraria, stockownership or options, expert testimony, grants or patents received orpending, or royalties.

References

Papers of special note have been highlighted as either of interest (•) or ofconsiderable interest (••) to readers.

1. Generic drugs. Ottawa (Canada): Canadian Agency for DrugTechnologies in Health (CADTH); 2012 [cited 2016 Sep 12].Available from: https://www.cadth.ca/generic-drugs

•• This summarizes the requirements for a second-entry productto be considered a generic. It includes important definitionsand concepts related to generics.

2. Generic drugs: questions and answers. Silver Spring (MD): U.S. Food& Drug Administration; 2015. [cited 2016 Sep 7]. Available from:

160 I. MAYERS AND M. BHUTANI

http://www.fda.gov/Drugs/ResourcesForYou/Consumers/QuestionsAnswers/ucm100100.htm

3. Health Canada. Guidance document: conduct and analysis of com-parative bioavailability studies. Ottawa: Health Canada; 2012.

4. Health Canada. Guidance document: comparative bioavailabilitystandards: formulations used for systemic effects. Ottawa: HealthCanada; 2012.

5. Health Canada. Draft guidance document: data requirements forsafety and effectiveness of subsequent market entry inhaled corti-costeroid products for use in the treatment of asthma. Ottawa:Health Canada; 2011.

•• This guidance outlines an approach to establishing BE usingchanges in inflammatory markers rather than relying uponchanges in lung function.

6. Government of Canada. Canadian food and drug regulationsc.08.002 and c.08.002.1. Ottawa: Minister of Justice, Canada; 2016.

7. Evans C, Cipolla D, Chesworth T, et al. Equivalence considerationsfor orally inhaled products for local action-isam/ipac-rs europeanworkshop report. J Aerosol Med Pulm Drug Deliv. 2012;25:117–139.

• This manuscript describes the EMA’s approach to in vitro test-ing for evaluating a second-entry product.

8. Government of Canada. Policy: interpretation of identical medicinalingredient. Ottawa: Health Canada; 2003.

9. Health Canada. Guidance to establish equivalence or relativepotency of safety and efficacy of a second entry short-actingbeta2-agonist metered dose inhaler. Ottawa: Health Canada; 1999.

10. Health Canada. Health Canada scientific advisory committee onrespiratory and allergy therapies - record of proceedings (item13). Ottawa: Health Canada; 2008.

11. Health Canada. Guidance for industry: pharmaceutical quality ofinhalation and nasal products. Ottawa: Health Canada; 2006.

12. Lee SL, Adams WP, Li BV, et al. In vitro considerations to supportbioequivalence of locally acting drugs in dry powder inhalers forlung diseases. AAPS J. 2009;11:414–423.

• This manuscript describes the limitations of utilizing PK stu-dies only in evaluating inhaled second-entry products.

13. Health Canada. Health Canada scientific advisory committee onrespiratory and allergy therapies - record of proceedings. Ottawa:Health Canada; 2009.

14. Science & research - critical path opportunities for generic drugs.Silver Spring (MD): U.S. Food and Drug Administration; 2007. [cited2016 Sep 8]. Available from: http://www.fda.gov/ScienceResearch/S p e c i a l T o p i c s / C r i t i c a l P a t h I n i t i a t i v e /CriticalPathOpportunitiesReports/ucm077250.htm

15. Horhota S. Running against the wind. Can the classic paradigm forbe really be extended to locally acting oindp? An analysis of thecurrent scientific and regulatory dilemmas of inhaler evaluation.Respiratory Drug Delivery. 2012;1:283–290.

16. Center for Drug Evaluation and Research. Draft guidance: bioavail-ability and bioequivalence studies for nasal aerosols and nasalsprays for local action. Rockville: Center for Drug Evaluation andResearch; 2003.

17. U.S. Food and Drug Administration. Draft guidance on bioequiva-lence of fluticasone propionate; salmeterol xinafoate. Silver Spring:U.S. Food and Drug Administration; 2013.

•• This guidance from the FDA outlines the multiple steps andthe complexity for a second-entry product to be accepted as ageneric for salmeterol/fluticasone inhaler.

18. U.S. Food and Drug Administration. Draft guidance on indacaterolmaleate. Silver Spring: U.S. Food and Drug Administration; 2016.

19. U.S. Food and Drug Administration. Draft guidance on ipratropiumbromide. Silver Spring: U.S. Food and Drug Administration; 2015.

20. U.S. Food and Drug Administration. Draft guidance on aclidiniumbromide. Silver Spring: U.S. Food and Drug Administration; 2015.

21. U.S. Food and Drug Administration. Draft guidance on beclometha-sone dipropionate. Silver Spring: U.S. Food and DrugAdministration; 2016.

22. U.S. Food and Drug Administration. Draft guidance on ciclesonide.Silver Spring: U.S. Food and Drug Administration; 2016.

23. U.S. Food and Drug Administration. Draft guidance on formoterolfumarate; mometasone furoate. Silver Spring: U.S. Food and DrugAdministration; 2016.

24. U.S. Food and Drug Administration. Draft guidance on budesonide;formoterol fumarate dihydrate. Silver Spring: U.S. Food and DrugAdministration; 2015.

25. U.S. Food and Drug Administration. Draft guidance on albuterolsulfate. Silver Spring: U.S. Food and Drug Administration; 2013.

26. U.S. Food and Drug Administration. Draft guidance on levalbuteroltartrate. Silver Spring: U.S. Food and Drug Administration; 2015.

27. European Medicines Agency. Guideline on the requirements forclinical documentation for orally inhaled products (OIP) includingthe requirements for demonstration of therapeutic equivalencebetween two inhaled products for use in the treatment of asthmaand chronic obstructive pulmonary disease (COPD) in adults and foruse in the treatment of asthma in children and adolescents. London:European Medicines Agency; 2009.

28. Adams WP, Ahrens RC, Chen ML, et al. Demonstrating bioequiva-lence of locally acting orally inhaled drug products (OIPS): workshopsummary report. J Aerosol Med Pulm Drug Deliv. 2010;23:1–29.

29. Fults K, Cyr TD, Hickey AJ. The influence of sampling chamberdimensions on aerosol particle size measurement by cascadeimpactor and twin impinger. J Pharm Pharmacol. 1991;43:726–728.

30. Dunbar C, Mitchell J. Analysis of cascade impactor mass distribu-tions. J Aerosol Med. 2005;18:439–451.

31. Apiou-Sbirlea G, Newman S, Fleming J, et al. Bioequivalence ofinhaled drugs: fundamentals, challenges and perspectives. TherDeliv. 2013;4:343–367.

32. De Backer J, Van Holsbeke C, Vos W, et al. Assessment of lungdeposition and analysis of the effect of fluticasone/salmeterolhydrofluoroalkane (HFA) pressurized metered dose inhaler (PMDI)in stable persistent asthma patients using functional respiratoryimaging. Expert Rev Respir Med. 2016;10:927–933.

33. Pritchard JN. The influence of lung deposition on clinical response.J Aerosol Med. 2001;14(Suppl 1):S19–26.

34. Scheuch G, Sommerer K, Meyer T, et al. Lung deposition studies:the challenge of interpretation. Respiratory Drug Delivery.2004;2:513–516.

35. Liu X, Jin L, Upham JW, et al. The development of models for theevaluation of pulmonary drug disposition. Expert Opin Drug MetabToxicol. 2013;9:487–505.

36. Brand P, Friemel I, Meyer T, et al. Total deposition of therapeuticparticles during spontaneous and controlled inhalations. J PharmSci. 2000;89:724–731.

37. Devadason SG, Chan HK, Haeussermann S, et al. Validation ofradiolabeling of drug formulations for aerosol deposition assess-ment of orally inhaled products. J Aerosol Med Pulm Drug Deliv.2012;25(Suppl 1):S6–9.

38. Fleming J, Conway J, Majoral C, et al. The use of combined singlephoton emission computed tomography and x-ray computedtomography to assess the fate of inhaled aerosol. J Aerosol MedPulm Drug Deliv. 2011;24:49–60.

39. Fleming JS, Conway JH. Three-dimensional imaging of aerosoldeposition. J Aerosol Med. 2001;14:147–153.

40. Chrystyn H. Methods to identify drug deposition in the lungsfollowing inhalation. Br J Clin Pharmacol. 2001;51:289–299.

41. Eberl S, Chan HK, Daviskas E. Spect imaging for radioaerosoldeposition and clearance studies. J Aerosol Med. 2006;19:8–20.

42. Hochrainer D, Hölz H, Kreher C, et al. Comparison of the aerosolvelocity and spray duration of respimat soft mist inhaler and pres-surized metered dose inhalers. J Aerosol Med. 2005;18:273–282.

43. Pauwels R, Newman S, Borgström L. Airway deposition and airwayeffects of antiasthma drugs delivered from metered-dose inhalers.Eur Respir J. 1997;10:2127–2138.

44. Newman S, Fleming J. Challenges in assessing regional distributionof inhaled drug in the human lungs. Expert Opin Drug Deliv.2011;8:841–855.

45. Harrison LI, Novak CC, Needham MJ, et al. Comparative pulmonaryfunction and pharmacokinetics of fluticasone propionate and

EXPERT OPINION ON DRUG DELIVERY 161

salmeterol xinafoate delivered by two dry powder inhalers topatients with asthma. J Aerosol Med Pulm Drug Deliv.2011;24:245–252.

46. Daley-Yates PT, Parkins DA, Thomas MJ, et al. Pharmacokinetic,pharmacodynamic, efficacy, and safety data from two randomized,double-blind studies in patients with asthma and an in vitro studycomparing two dry-powder inhalers delivering a combination ofsalmeterol 50 μg and fluticasone propionate 250 μg: implicationsfor establishing bioequivalence of inhaled products. Clin Ther.2009;31:370–385.

47. Talton J, Fitz-Gerald J, Singh R, et al. Nano-thin coatings forimproved lung targeting of glucocorticoid dry powders: in-vitroand in-vivo characteristics. Respiratory Drug Delivery. 2000;1:67–74.

48. Hohlfeld JM, Sharma A, van Noord JA, et al. Pharmacokinetics andpharmacodynamics of tiotropium solution and tiotropium powderin chronic obstructive pulmonary disease. J Clin Pharmacol.2014;54:405–414.

49. Crapo RO, Casaburi R, Coates AL, et al. Guidelines for methacholineand exercise challenge testing. Am J Respir Crit Care Med.2000;161:309–329.

50. Clearie KL, Williamson PA, Meldrum K, et al. Pharmacokinetic andpharmacodynamic comparison of hydrofluoroalkane and chloro-fluorocarbon formulations of budesonide. Br J Clin Pharmacol.2011;71:504–513.

51. Horhota ST, van Noord JA, Verkleij CB, et al. In vitro, pharmacoki-netic, pharmacodynamic, and safety comparisons of single andcombined administration of tiotropium and salmeterol in copdpatients using different dry powder inhalers. AAPS J.2015;17:871–880.

52. Kuna P, Govoni M, Lucci G, et al. Pharmacokinetics and pharmaco-dynamics of an extrafine fixed pmdi combination of beclometa-sone dipropionate/formoterol fumarate in adolescent asthma. Br JClin Pharmacol. 2015;80:569–580.

53. Reddel HK, Taylor DR, Bateman ED, et al. An official AmericanThoracic Society/European Respiratory Society statement: asthmacontrol and exacerbations: standardizing endpoints for clinicalasthma trials and clinical practice. Am J Respir Crit Care Med.2009;180:59–99.

54. Wise RA, Anzueto A, Calverley P, et al. The tiotropium safety andperformance in respimat trial (tiospir), a large scale, randomized, con-trolled, parallel-group trial-design and rationale. Respir Res. 2013;14:40.

55. Lavorini F, Geri P, Camiciottoli G, et al. Agreement between twomethods for assessing bioequivalence of inhaled salbutamol. PulmPharmacol Ther. 2008;21:380–384.

56. Lavorini F, Geri P, Mariani L, et al. Speed of onset of bronchodilatorresponse to salbutamol inhaled via different devices in asthmatics:a bioassay based on functional antagonism. Br J Clin Pharmacol.2006;62:403–411.

57. Sharma A, Kerstjens HA, Aalbers R, et al. Pharmacokinetics oftiotropium administered by respimat(r) in asthma patients: analysisof pooled data from phase II and III clinical trials. Pulm PharmacolTher. 2017;42:25–32.

58. Pirozynski M, Sosnowski TR. Inhalation devices: from basic scienceto practical use, innovative vs. generic products. Expert Opin DrugDeliv. 2016;13:1559–1571.

59. Lavorini F, Braido F, Baiardini I, et al. Asthma and copd: interchange-able use of inhalers. A document of Italian Society of Allergy, Asthmaand Clinical Immmunology (SIAAIC) & Italian Society of RespiratoryMedicine (SIMER). Pulm Pharmacol Ther. 2015;34:25–30.

60. Molimard M, Raherison C, Lignot S, et al. Assessment of handling ofinhaler devices in real life: an observational study in 3811 patientsin primary care. J Aerosol Med. 2003;16:249–254.

61. Thomas M, Price D, Chrystyn H, et al. Inhaled corticosteroids forasthma: impact of practice level device switching on asthma con-trol. BMC Pulm Med. 2009;9:1.

62. Braido F, Lavorini F, Blasi F, et al. Switching treatments in COPD:implications for costs and treatment adherence. Int J ChronObstruct Pulmon Dis. 2015;10:2601–2608.

63. Wieshammer S, Dreyhaupt J. Dry powder inhalers: which factors deter-mine the frequency of handling errors? Respiration. 2008;75:18–25.

64. Melani AS, Zanchetta D, Barbato N, et al. Inhalation technique andvariables associated with misuse of conventional metered-doseinhalers and newer dry powder inhalers in experienced adults.Ann Allergy Asthma Immunol. 2004;93:439–446.

65. Leiner S, Parkins D, Lastow O. Inhalation devices and patient inter-face: human factors. AAPS J. 2015;17:457–461.

162 I. MAYERS AND M. BHUTANI