<%BANNER%>

Bioequivalence of Inhaled Corticosteroids

Permanent Link: http://ufdc.ufl.edu/UFE0024287/00001

Material Information

Title: Bioequivalence of Inhaled Corticosteroids A Pharmacokinetic Approach
Physical Description: 1 online resource (128 p.)
Language: english
Creator: Goyal, Navinkumar
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: bioequivalence, cascade, clinical, corticosteroids, exposure, glucocorticoids, guidelines, inhaled, inhalers, invitro, invivo, modeling, pharmacokinetics, recommendations, simulation, steroids, study
Pharmacy -- Dissertations, Academic -- UF
Genre: Pharmaceutical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Inhaled drug products are the treatment of choice for respiratory conditions like asthma and chronic obstructive pulmonary disease. Since the patents for several inhaled corticosteroids (ICS) expired, less expensive generic inhalers are possible. Inhalers are intended to provide drug for local lung delivery and hence pose special considerations while conducting bioequivalence (BE) studies. The generic inhaler needs to demonstrate equivalent efficacy and safety as that of the original inhaler. Lack of consensus exists between industry, academia and regulatory authorities over criteria to establish BE of inhaled drugs. The US Food and Drug Administration recommends use of in vitro studies for drug product performance, a clinical efficacy (Pharmacodynamic) study to ensure equivalence of local effects and a systemic exposure (Pharmacokinetic) study to ensure similar safety profiles. However, pharmacodynamic studies have proven to be an inefficient tool to evaluate bioequivalence due to the poor dose-response relationship of ICS. Any alternative approach to the current FDA strategy must address the following three key questions to ensure BE of inhalation drugs ? How much drug is available in the lung? How much is absorbed orally (only relevant for drugs with significant oral bioavailability)? ? Where is the drug deposited (central vs. peripheral lung)? ? How fast is the drug absorbed? We hypothesize that a pharmacokinetic (PK) approach that utilizes PK endpoints such as area under the curve (AUC) and Cmax, could answer these questions. Population pharmacokinetic simulations were used to test this hypothesis. Simulations demonstrated that for ICS with negligible oral bioavailability, the AUC will be able to detect a difference ( > 20%) in the amount of ICS available to the lung when 30-50 subjects are used. For slowly dissolving drugs, the AUC is also sensitive to differences in the regional lung deposition (central to peripheral (C/P) deposition ratio). Similar AUCs may result from a lower lung dose with more peripheral deposition and higher dose with more central deposition for slowly dissolving drugs. To ensure that a similar AUC is not due to the simultaneous difference in these two parameters, one may use in vitro data from cascade impactor studies or conduct PK studies in asthmatics to support the BE decisions in such scenarios. The PK endpoint - Cmax can detect differences in the absorption rate of the drug deposited from two inhalers. To summarize, the PK approach provides a cost and time effective tool to evaluate the bioequivalence of inhaled corticosteroids.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Navinkumar Goyal.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Hochhaus, Guenther.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024287:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024287/00001

Material Information

Title: Bioequivalence of Inhaled Corticosteroids A Pharmacokinetic Approach
Physical Description: 1 online resource (128 p.)
Language: english
Creator: Goyal, Navinkumar
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: bioequivalence, cascade, clinical, corticosteroids, exposure, glucocorticoids, guidelines, inhaled, inhalers, invitro, invivo, modeling, pharmacokinetics, recommendations, simulation, steroids, study
Pharmacy -- Dissertations, Academic -- UF
Genre: Pharmaceutical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Inhaled drug products are the treatment of choice for respiratory conditions like asthma and chronic obstructive pulmonary disease. Since the patents for several inhaled corticosteroids (ICS) expired, less expensive generic inhalers are possible. Inhalers are intended to provide drug for local lung delivery and hence pose special considerations while conducting bioequivalence (BE) studies. The generic inhaler needs to demonstrate equivalent efficacy and safety as that of the original inhaler. Lack of consensus exists between industry, academia and regulatory authorities over criteria to establish BE of inhaled drugs. The US Food and Drug Administration recommends use of in vitro studies for drug product performance, a clinical efficacy (Pharmacodynamic) study to ensure equivalence of local effects and a systemic exposure (Pharmacokinetic) study to ensure similar safety profiles. However, pharmacodynamic studies have proven to be an inefficient tool to evaluate bioequivalence due to the poor dose-response relationship of ICS. Any alternative approach to the current FDA strategy must address the following three key questions to ensure BE of inhalation drugs ? How much drug is available in the lung? How much is absorbed orally (only relevant for drugs with significant oral bioavailability)? ? Where is the drug deposited (central vs. peripheral lung)? ? How fast is the drug absorbed? We hypothesize that a pharmacokinetic (PK) approach that utilizes PK endpoints such as area under the curve (AUC) and Cmax, could answer these questions. Population pharmacokinetic simulations were used to test this hypothesis. Simulations demonstrated that for ICS with negligible oral bioavailability, the AUC will be able to detect a difference ( > 20%) in the amount of ICS available to the lung when 30-50 subjects are used. For slowly dissolving drugs, the AUC is also sensitive to differences in the regional lung deposition (central to peripheral (C/P) deposition ratio). Similar AUCs may result from a lower lung dose with more peripheral deposition and higher dose with more central deposition for slowly dissolving drugs. To ensure that a similar AUC is not due to the simultaneous difference in these two parameters, one may use in vitro data from cascade impactor studies or conduct PK studies in asthmatics to support the BE decisions in such scenarios. The PK endpoint - Cmax can detect differences in the absorption rate of the drug deposited from two inhalers. To summarize, the PK approach provides a cost and time effective tool to evaluate the bioequivalence of inhaled corticosteroids.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Navinkumar Goyal.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Hochhaus, Guenther.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024287:00001


This item has the following downloads:


Full Text

PAGE 1

BIOEQUIVALENCE OF INHALED CORTICOSTEROIDS A PHARMACOKINETIC APPROACH By NAVINKUMAR S. GOYAL A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Navinkumar S. Goyal 2

PAGE 3

To my mom 3

PAGE 4

ACKNOWLEDGMENTS I would like to acknowledge Dr. Guenther Ho chhaus, my supervisor, for the opportunity to pursue my graduate studies in his lab. I am gr ateful for the help, guidance and support I have received from him all these year s. He has been a great mentor. I am thankful to Dr. Jeffrey Hughes, Dr. Hartmut Derendorf and Dr. Andre Mauderli for being a part of my graduate committee. I am also grateful to my labmat es and other graduate students in the department for all their support. I would also like to thank Dr. Sreedharan Sabarinath, Dr. Varun Goel and Dr Jian Xu for the great technical arguments and discussions we have had through my project work. I cannot thank enough my friends Dr. Preeti Ya dava, Vineet Miharia, Hemal Vyas and Naresh Pai who made have been a great company all these years during my graduate studies and provided a home away from home. Finally, I would like to thank my family, esp ecially my sister Dr. Meena Goyal for their love and encouragement without which I would have never ever been able to accomplish any of this. I am thankful to my grandfather Ramkumar for his everlasting support and blessings that have helped me through all the tough times. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ..........7LIST OF FIGURES.........................................................................................................................8ABSTRACT.....................................................................................................................................9CHAPTER 1 AIRWAYS AND ASTHMA..................................................................................................11Introduction................................................................................................................... ..........11Airways and Lung Function...................................................................................................11Mucociliary Clearance............................................................................................................12Asthma Pathophysiology........................................................................................................152 ASTHMA THERAPY............................................................................................................18Inhaled Corticosteroids...........................................................................................................18Fate of Inhaled Corticosteroids...............................................................................................20Inhalers...................................................................................................................................21Issues Involved in Bioequivalence Testing............................................................................243 POTENTIAL TOOLS............................................................................................................27Introduction................................................................................................................... ..........27Potential tools................................................................................................................ .........27Pharmacodynamic Studies......................................................................................................28Pharmacokinetic (PK) Studies................................................................................................34Scintigraphic Studies.......................................................................................................... ....37Correlation of Lung Deposition Studies.................................................................................38In vitro Studies........................................................................................................................40Summary.................................................................................................................................424 REGULATORY RECOMMENDATIONS............................................................................46Health Canada.........................................................................................................................46European Medicines Agency..................................................................................................48United Kingdom.....................................................................................................................50United States Food and Drug Administration........................................................................51Regulatory Approach Comparisons........................................................................................52Summary.................................................................................................................................53Conclusion..............................................................................................................................54 5

PAGE 6

Hypothesis..............................................................................................................................555 HOW MUCH DRUG IS AVAILABLE?...............................................................................60Introduction:...........................................................................................................................60Methods..................................................................................................................................62Simulations.................................................................................................................... .........63Results.....................................................................................................................................64Conclusion..............................................................................................................................66Discussion...............................................................................................................................666 WHERE DOES THE DRUG DEPOSIT?..............................................................................75Introduction................................................................................................................... ..........75Absorption and Muco-ciliary Clearance.................................................................................76Methods..................................................................................................................................79Simulations.................................................................................................................... .........80Results.....................................................................................................................................81Special Scenario............................................................................................................... .......82Conclusion..............................................................................................................................84Discussion...............................................................................................................................857 HOW LONG DOES DRUG STAY IN LUNG?....................................................................92Introduction................................................................................................................... ..........92Methods and Simulations.......................................................................................................9 4Results.....................................................................................................................................95Conclusion..............................................................................................................................96Discussion...............................................................................................................................968 STUDY DESIGN................................................................................................................. 100Introduction................................................................................................................... ........100Bioequivalence study design................................................................................................101APPENDIX A NONMEM CODE FOR PK MODE L OF FP AFTER INHALATION...............................106B R CODE FOR AUC A ND CMAX CALCULATION.........................................................108LIST OF REFERENCES.............................................................................................................110BIOGRAPHICAL SKETCH.......................................................................................................128 6

PAGE 7

LIST OF TABLES Table page 2-1 Pharmacokinetic properties of inhaled corticosteroids......................................................253-1 Summary of few studies with different dose levels...........................................................444-1 Clinical models to determine relative potency of ICS.......................................................565-1 Parameters and their associated variability........................................................................695-2 Results from simulations....................................................................................................706-1 Variability on various paramete rs in the simulated POPPK model...................................866-2 Simulation results with Kmuc~0.5/hr Same dose available through reference and generic inhaler. Slowly di ssolving drugs such as FP.........................................................876-3 Simulation results with Kmuc~0.2/hr Same dose available through reference and generic inhaler. Slowly di ssolving drugs such as FP.........................................................886-4 Simulation results for special scenari o. Change in the drug dose and regional deposition pattern. Slowly di ssolving drugs such as FP....................................................897-1 Simulation results for inhalers with diffe rent absorption profiles. Slowly dissolving drugs such as FP.................................................................................................................98 7

PAGE 8

LIST OF FIGURES Figure page 1-1 The bronchial tree (Weibel model)....................................................................................171-2 Different processes of dissolution, absorp tion and muco-ciliary clearance occurring on particles deposited in lung post inhalation....................................................................172-1 Activation of anti-inflammatory gene e xpression by corticosteroids (Barnes [30])..........262-2 Fate of inhaled corticosteroid............................................................................................. 263-1 Ahrens design of crossover clinical study for BE of ICS..................................................454-2 Health Canada Draft Guidance Requirements for ICS Bioequivalence..........................584-3 EMEA guideline scheme for bioequivalence....................................................................595-1 Flow chart for modeling and simulation of trials...............................................................715-2 Goodness of fit plots. The open circles are the observe d values and the lines are model fits..................................................................................................................... ......725-3 Individual and population fits from the model..................................................................735-4 Spaghetti plots of plasma concentrations from simulations..............................................746-1 Different central and peripheral de position in healthy and asthmatics..............................906-2 Compartmental model for inhaled flu ticasone propionate used for simulation.................917-1 Compartmental model scheme for inhale d and intravenous fluticasone propionate used for simulation.............................................................................................................99 8

PAGE 9

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy BIOEQUIVALENCE OF INHALED CORTICOSTEROIDS A PHARMACOKINETIC APPROACH By Navinkumar S. Goyal August 2009 Chair: Guenther Hochhaus Major: Pharmaceutical Sciences Inhaled drug products are the treatment of c hoice for respiratory conditions like asthma and chronic obstructive pulmonary disease. Since the patents for several inhaled corticosteroids (ICS) expired, less expensive generi c inhalers are possible. Inhale rs are intended to provide drug for local lung delivery and hence pose special considerations while c onducting bioequivalence (BE) studies. The generic inhaler needs to demonstr ate equivalent efficacy and safety as that of the original inhaler. Lack of consensus ex ists between industry, academia and regulatory authorities over criteria to es tablish BE of inhaled drugs. The US Food and Drug Administration recommends use of in vitro studies for drug product performance a clinical efficacy (Pharmacodynamic) study to ensure equivalence of local effects and a systemic exposure (Pharmacokinetic ) study to ensure sim ilar safety profiles. However, pharmacodynamic studies have proven to be an inefficient tool to evaluate bioequivalence due to the poor dos e-response relationship of ICS. Any alternative approach to the current FDA strategy must address the follo wing three key questions to ensure BE of inhalation drugs 9

PAGE 10

10 How much drug is available in the lung? How mu ch is absorbed orally (only relevant for drugs with significant oral bioavailability)? Where is the drug deposited (c entral vs. peripheral lung)? How fast is the drug absorbed? We hypothesize that a pharmacoki netic (PK) approach that u tilizes PK endpoints such as area under the curve (AUC) and Cmax, could answer these questions. Population pharmacokinetic simulations were used to test this hypothesis. Simulations de monstrated that for ICS with negligible oral bioavailability, the AUC will be able to detect a difference (>20%) in the amount of ICS available to the lung when 30-50 subjects are used. For slowly dissolving drugs, the AUC is also sensitive to differences in the regional lung deposition (central to peripheral (C/P) deposition ratio). Similar AUCs may result fr om a lower lung dose w ith more peripheral deposition and higher dose with more central de position for slowly dissolving drugs. To ensure that a similar AUC is not due to the simultane ous difference in these two parameters, one may use in vitro data from cascade impactor studies or conduct PK studies in asthmatics to support the BE decisions in such scenarios. The PK endpoint Cmax can detect differences in the absorption rate of the drug deposited from two inhalers. To summarize, the PK approach provides a cost and time effective tool to evaluate the bioequivalen ce of inhaled corticosteroids.

PAGE 11

CHAPTER 1 AIRWAYS AND ASTHMA Introduction The airways are an essential organ system that allows air to come in close contact of the blood. They also present a unique system for sy stemic as well as local drug delivery. This chapter explains the structural details of the lung. We discus s pulmonary processes such as mucociliary clearance that occur in the lung and its significance in drug delivery. Asthma and its pathophysiology are also discussed briefly to get a better insight into the disease process. Airways and Lung Function The airways originate at trachea, branch in to smaller and smaller airway passages and terminate in the alveolar sacs that could be de scribed as a pulmonary tree. The structure can be well explained with the help of figure 1-1 Th e tree trunk is represented by the trachea which bifurcates into the two main bronchi. The br onchi further subdivide into smaller bronchi, bronchioles and terminal bronchiol es. This is followed by the re spiratory bronchi oles, alveolar ducts and ultimately the alveolar sacs. The Weibel model categorizes the ai rway into 24 distinct generations with trachea as the 0th generation and the alveolar sacs as 23rd generation [1]. As one moves from the upper airways or trachea and bron chi towards the lower airways, the airway caliber decreases. There are functional differences with the upper airways acting as conducting airways and the lower airways as respiratory airways where actual gas exchange occurs. Apart from gas exchange, the lung performs other impo rtant functions such as maintaining the acid base balance and consequently blood pH. Certai n pulmonary cells secret e substances such as histamine, prostaglandins and le ukotrienes that exert local and/ or systemic action. The enzymes produced by the pulmonary epithelium are involve d in metabolism of selective substances that may be of exogenous or endogenous nature [2]. 11

PAGE 12

The lung epithelium is the cell sheet lining th e airway lumen. It se parates the internal environment of the lung from the external enviro nment and is exposed to substances such as gases, aerosols along with any pa rticulate matter that may be inha led. The different types of cells that line the airway epithelium confer various functions to the airway epithelium. These functions depend on the area where these cell type s may be located. Some of these cell types include ciliated columnar cells, mucous secreting goblet cells, basal cells, dendrite cells, alveolar type I and II cells, basal and serous cells. Th e ciliated cells extend from the trachea to the terminal bronchioles. They bathe in the flui d bilayer secreted by the goblet cells. The fluid bilayer contains a sol or watery phase adjace nt to the epithelium and a gel or mucus phase adjacent to the lumen. Together they constitu te what is called the mucociliary clearance processes which provide an important defensive mechanism of the lung. Mucociliary Clearance The fluid bilayer performs 4 important functions of: particle entrapment humidifying the inhaled air protecting the epithelium fr om being dehydrated and housing various antibacterial enzymes and proteins that prevent microbial colonization of the airways [3]. The ciliary processes beat ra pidly in coordination and push the mucus layer towards the pharynx out of the lung. This mucus may be expe ctorated or swallowed. Thus the mucociliary processes act as a defense mechanism by entrapping all xenobiotics. The mucus is propelled at different rates in different regions of the lung. Th e rate decreases towards the peripheral airways. The particle entrapment is of importance for as thma treatment since any drug particles that may be delivered to the lung will also be trapped in th is mucus layer. The drug particle has to dissolve in the lung fluids before it can in teract with receptor targets in the epithelium and be absorbed. This process will have to compete with the mucociliary clearance mechanism. 12

PAGE 13

The ciliated and mucus producing airway struct ures start disappeari ng as the bronchioles branch down further into terminal bronchioles a nd alveolar regions that are non-ciliated [4]. Consequently, the inhaled particle s may be retained in the alveolar region for long periods of time (in days). The rate of absorp tion in the upper or central airways is small as compared to that in the lower or peripheral regions. At the same time the absorption from the alveolar regions is very high. Studies have shown that the absorption rate from the peripheral compartment is twice as high as that from the central region [5, 6]. The inhaled drug particles (which are soluble in nature) will be absorbed befo re any significant clearance thro ugh mucociliary mechanism occurs in these peripheral regions of th e lung. The rates of dissolutio n, absorption and muco-ciliary clearance thus govern the amount of drug in the lung. This is well e xplained with figure 1-2 that has been adapted from Edsbacker et al. [7] It can be seen that some particles in the ciliated airways in the central region of the lung dissolve in the lung fl uids (mucus layer) and are absorbed, while some particles are cleared by th e ciliated cells before they are absorbed. The mucociliary clearance rates differ depending on the lung region [8]. Coughing may increase the mucus propulsion rate, whereas some other clinical conditions may slow down this process. The mucociliary process may also be a ffected (slower) in subjec ts with lung diseases [9]. For example in cystic fibrosis or bronchitis, there is hyper secretion of the mucus and ciliary dysfunction which leads to conges tion. It has been rightly men tioned that determination of mucociliary clearance is diffi cult to quantify [10] and depe nds upon nature of particles depositedsoluble or insoluble particles. Currently prescribed inhalation drugs can be considered to behave as soluble particles. The inhaled cort icosteroids act at the receptor level at intra or extracellular locations in the lung. Alternatively, the ICS may also be retained in interstitial fluid before being ultimately absorbed into systemic circulation or be cleared by mucociliary 13

PAGE 14

mechanism [11]. The ciliary processes are predom inantly active in the tracheo-bronchial region and such a phenomenon is absent in the peripheral region. Studies in the past have attempted to estimate the mucociliary clearance rates using radiolabeled particles such as liposom es, ferrous oxide particles, carbon particles, Teflon particles as well as drug molecules. These stud ies have found that the tracer pa rticles have a half life ranging from 0.5 hours to 24 hours depending on whether th ey are deposited centrally or peripherally within the lung [12-16]. It may take even days in some cases for the drug deposited in deep areas of the lung to be cleared. However, since th e ICS are soluble and absorbable drugs, the drug deposited in the peripheral area of the lung will be absorbed before being cleared over such a long period of time. Hence, only th e rapid muco-ciliary clearance which occurs from the central region of the lung would predominantly affect the amount of drug available for systemic absorption. Since some drug fraction may be cleare d by mucociliary transport before it is able to elicit its effect, this might reduce the dose availa ble to the lung. Any c ondition that changes the mucus secretion or clearance will affect the amou nt of drug available. Apart from this the drug characteristics such as lipophilicity may also in fluence the amount of dr ug available to the lung. For highly lipophilic and slowly dissolving ICS like fluticasone, the mucociliary clearance will affect the lung dose. This may not be the case for the more hydrophilic drug budesonide. We will discuss in the later chapters how the regional differences in the absorption rates and mucociliary clearances can be taken advantage of to understa nd and estimate the regional drug deposition in the lung. The lung function can be evaluated by measur ing various pulmonary volumes and airway caliber using spirometry that measures the volume of air entering and exiting the airways. These volumes and airway caliber change due to diseas e conditions. Various lung volumes include tidal 14

PAGE 15

volume, inspiratory reserve volume, expiratory reserve volume, residua l volume and total lung capacity. The peak expiratory flow rate, for ced expiratory flow m easures are noninvasive measures of airway caliber. So me of the important and commonl y used measurements include the FEV1 or the volume expired in first second and FVC or the forced vital capacity. The FEV can be normalized for various physiological factors such as sex, age and body weight. This enables comparison of lung functi on values to normal or expect ed estimates. For example the ratio of FEV1 and FVC in individuals with nor mal lung function is around 0.8 and in subjects with airway obstruction is lower than 0.8 [2]. Other highly speci alized equipments as well as invasive techniques are also available to measure airway resistance and dynamic lung compliance. This includes wholebo dy plethysmograph and pneumotachygraph. Asthma Pathophysiology Asthma and chronic obstructive pulmonary disorder (COPD) are chronic inflammatory diseases affecting the airways. There is no cure for these airway disorders, although they can be controlled so as to have fewer symptoms. Asthma is characterized by variable airflow limitation and airway hyper responsiveness [17, 18]. The asth ma symptoms or clinical features include wheezing (a whistling sound when one breathes ), coughing, chest tightness and troubled breathing, especially late night a nd early morning [19]. COPD cons titutes obstruc tion to airflow resulting in interference with breathing. COPD refers to bron chitis and emphysema, conditions that frequently coexist. [19]. Asthma results in occlusions of airway lumen by tenacious mucus secretions, infiltration of eosinophi ls and lymphocytes in the airway wall as well as shedding of airway epithelia [20]. There is also bronchial smooth muscle enlargement and thickening of reticular basement membrane in asthma [21]. The complex pathophysiology of asthma can be accounted for by a number of inflammatory cel ls which include mast cell, macrophages, dendritic cells, eosinophils and T lymphocytes. There is lack of sufficient evidence about the role 15

PAGE 16

of neutrophils in asthma. The airway structural cells such as the epithel ial cells, fibroblast and smooth muscle cells are important source of inflammatory mediator s such as cytokines [22]. The epithelial cells are also a ttributed to play an important role in the airway inflammatory response by interacting with the inhaled environmental sign als. Consequently they are a major target for the therapy with the inhaled corticosteroids [23] A number of inflammatory mediators have been implicated in asthma. These include cytoki nes, histamine, prostaglandins, leukotrienes, bradykinins, nitric oxide, endothe lins, growth factors. These mediators bring about airway smooth muscle contraction, increased mucus secr etion. Every mediator has several effects and thus estimation of contribution by any single mediator is difficult. It was initially thought that asthma was a disease of the central airways. Bronchoscopic studies have clearly shown that asthma is associ ated with infiltration of the major airways with chronic inflammatory cells, particularly lymphoc ytes and eosinophils [24]. There also exists upregulation of cytokine profiles. Postmortem systemic studies have confirmed the presence of inflammation throughout the airways [25, 26]. Advances in immunohistochemical techniques have also provided evidence that the peripheral ai rways are significant sites of inflammation in asthma [24, 27]. Comparison betw een several markers of infl ammation in mucosal biopsy specimens and BAL fluid from healthy control s ubjects, patients with intermittent asthma, and patients with mild-to moderate persistent asth ma found significant increase in the alveolar macrophage activation in patients with asthma comp ared with control patie nt [28]. All this has led to an increased acceptance that asthma is an inflammatory disease of lower and upper airways. Consequently any asthma therapy should be targeted towards the entire lung rather than just the central airways. 16

PAGE 17

Figure 1-1. The bronchial tree (Weibel model) Figure 1-2. Different processes of dissolution, absorption and muco -ciliary clearance occurring on particles deposited in lung post inhalation 17

PAGE 18

CHAPTER 2 ASTHMA THERAPY The rationale for asthma therapy has evolve d with better understanding of the underlying inflammatory processes. Rather than solely relying on relieving symptoms there has been increased use of agents that control underlying inflammation. Asthma therapy could be classified into mainly anti-inflammatory or controller ag ents (corticosteroids and cromolyn sodium) and bronchodilators ( 2 agonists) or relievers. There has also been some degree of success reported with the anti-leukotriene drugs. This is especially attributed to the fact that corticosteroids have been unable to effectively block the leukotriene synthesis. Inhaled Corticosteroids Inhaled corticosteroids (ICS) represent one of the most effective anti-inflammatory treatment options by acting on a variety of targets such as eosinophils, macrophages, T lymphocytes, dendritic cells, mast cells as well as the structural epithelial cells. Corticosteroids reduce the recruitment of infl ammatory cells into the airw ay. ICS do so by suppressing the production of chemotactic mediators and adhesion molecules. ICS also inhibit the survival of eosinophils, T lymphocytes and mast cells in the airways [29]. This poten tially blocks a major source of inflammatory mediators in asthma. The process begins with the ICS binding to the glucocorticoids receptors in the cytoplasm of target cells. Binding of corticosteroids to GR results in a conformational change which allows the ICS-GR complex to translocate to the nucleus or interact with cy toplasmic transcription factors. Once in the nucleus, the GR complex bi nds as a dimer to specific DNA sites, specific nucleotide palindromic sequences termed gluco corticoid response elements (GRE). Then the transcription of specific genes can be increased (t ransactivation) or decrea sed (transrepression). This depends on whether the GRE is positive or ne gative. Two GR molecules bind together as a 18

PAGE 19

homodimer and bind to GRE. This brings about increase in gene transcription, also known as transactivation. The binding of GR homodimer to the negative GRE results in cis-repression leading to gene suppression. It is estimated that between 10 a nd 100 genes per cell are directly regulated by the ICS. Many genes are regulated indirectly through th e interaction between corticosteroids and other transcription factors and co activ ators. CREB (cyclic adenosine monophosphate response elementbinding protein) (CBP), p300/CBP associated factor (PCAF) or steroid receptor coactivator-1 (SRC-1) are such coactivator molecules. In particular, transactivation by corticosteroids increases the expr ession of anti-inflammatory proteins such as annexin-1 (lipocortin-1), inte rleukin-1 receptor antagonist (IL -1ra), interleukin-10 (IL-10), secretory leukocyte inhibitory protein (SLPI), ne utral endopeptidase and the inhibitory protein (I B) of nuclear factorB (NFB). Corticosteroids switch on the synthesis of two proteins that affect the signal transduction pathways, glucocorticoid induced leucine zipper protein (GILZ) and MAP kinase phosphatase-1 (MKP-1) [30]. Cortic osteroids also inhibit the synthesis of many inflammatory proteins through suppression of genes that encode them. In spite of lacking GRE in their promoter regions, many inflammatory ge nes activated in asthma are repressed by corticosteroids. The effects of pro-inflammatory transcription factors such as AP-1 and NF-kB are inhibited by corticosteroids [31, 32]. Cor ticosteroids are also able to enhance the transcription of th e gene encoding the 2-adrenergic receptor. This is postulated to result in reversing and/or preventing the down-regulation of the 2 receptors possib ly induced by longterm treatments with 2-agonist bronchodilators [33]. T hus, corticosteroids reduce airway inflammation and hyperresponsiveness by altering the production of inflammatory mediators. Figure 2-2 explains this complex mechanism of action. The decrease in inflammation can be clinically correlated to imp roved asthma symptoms. 19

PAGE 20

The negative GRE examples include genes that regulate HPA axis (pro-opiomelanocortin and corticotrophin releasi ng factor), bone metabolism (osteocalci n) and skin structure (keratins) [29]. The exact molecular mechanisms of costicosteroid induced side effects is not clear [29]. It has been thought that gene activ ation may be the cause. Few system ic side effects with ICS are osteoporosis, reduced growth velocity in children, sk in thinning, cataracts and glaucoma. Exogenous corticosteroids bring about HPA axis suppression by downregulating the adrenocorticotropic hormone (ACTH) production by the same feedback inhibition loops that control endogenous glucocorticoid production. This leads to adrenal s uppression and decreased cortisol levels. ICS dose, duration of treatment and time of drug administration are factors shown to affect the degree of adrenal suppression [34]. Fate of Inhaled Corticosteroids Considerable development has taken place on the drug as well as the inhaler device development fronts. This has had significant effects on pulmonary drug targeting, since pulmonary targeting is a function of drug as well as inhaler device. Before proceeding to discuss the various inhaler devices lets understand the dr ug delivery process through inhalers. The fate of the inhaled drug can be explaine d well by the figure 2-3. A major fraction (40-90%) of the dose from the inhaler gets deposited in the upper resp iratory tract while a part of it (10-60%) reaches the lungs. The fraction of the drug deposited in the oropharyngeal region gets swallowed and enters the systemic circulat ion via GI absorption [35]. Hence, the total systemic bioavailability of an inhaled glucocorticoid is the sum of the oral and pulmonary bioavailable faction. The free fraction of ICS in the systemic circulation binds to the systemic glucocorticoids receptors to produce the systemic side effects. The drug is eventually eliminated from the systemic circulation mainly by hepatic clearance mechanisms. An ideal ICS is a compound with high pulmonary activity which is inactivated rapidly and efficiently after absorp tion. Such an agent would have high pulmonary 20

PAGE 21

activity but minor or no systemic side effects when administered in the desired therapeutic dose range [36]. At the same time the oral bioavailabil ity should be negligible so that the dose that enters the GI tract does not enter the systemic circulation. FP and BUD are two such ICS with favorable pulmonary effects and minimal system ic side effects. The newer ICS such as fluticasone propionate, mometasone furoate and ciclesonide have ne gligible oral bioavailability (<1%) as compared to the older ones such a s beclomethasone and budesonide (10-30%) [37]. These drugs also have high pr otein binding and high systemic clearance. This reduces the systemic side effects of the drug that is abso rbed via lung. However, an ICS with high plasma protein binding will also possess high tissue binding. This will result in lower desired effects due to high lung tissue binding. The pharmacokinetic prope rties of various ICS ar e presented in table 2-1. After deposition in the lung, the drug dissolves in the lung fluids, interacts with the receptors and is finally absorbed into the systemic circul ation. A fraction of the undi ssolved drug particles may be cleared out of the lung by the mucocilia ry transport processes as discussed in the preceding section. In a pharmacokinetic study, the dr ug levels are measured in plasma which is downstream from where the drug is deposited. Ther e is a general view that the systemic drug levels do not contain sufficient information about drug fate in the lung. We hypothesize that the systemic drug levels carry sufficient informa tion about what goes on with the drug in the lung. The use of these systemic drug levels in evalua ting bioequivalence will be discussed later. Inhalers Devices available for delivery of the ICS in clude pressurized metered dose inhaler (pMDI), dry powder inhaler (DPI) and nebul izers. Inhaler development, necessary to switch from the CFC propellants to the HFA propellants also re sulted in an improved design. The newer inhaler designs have allowed the formation of smaller droplet sizes and increa sed pulmonary deposition. The HFA propellants provide with an eco friendl y option as well as smaller particle size 21

PAGE 22

distribution (PSD) of the aerosol generated. The smaller PSD lead s to greater overall deposition (up to 60%) in addition to more peripheral de position. There are numerous studies comparing the efficiency of these devices in terms of amount of lung deposition. However, the difficulties in optimally handling pMDIs are well documented [38, 39]. These require sufficient coordination on part of the subject during actuation and inhaltion. The problems in co-coordinating actuation and inhalation from a pMDI can be minimized by placing a spacer device between the actuator and mouth. At the same time there is also greater pulmonary drug deposition with a spacer. There are studies describing the lung deposition of ICS from pMDIs with such spacer devices [40]. DPIs, on the other hand are gaining popularity and have been successfully shown to be better or at least equal in performance when co mpared to pMDIs [41-43] The drug is released when the patient inhales. Hence less coordination is required by the patient while using a DPI. Some DPIs provide drug alone or in combination with a carrier substance usually lactose. The drug may be available in pre fill ed blisters or capsules or as a reservoir system. They are available as single dose inhalers or multiple dos e inhalers as well. However, the DPIs are inspiratory flow driven and hence require the subj ect to have a forceful deep inhalation. This allows the breaking up of micron sized drug and diluent particle aggregates into smaller respirable particles in the oropharynx and the la rger airways which can then enter the lung. The higher the inspiratory flow, the smaller will be the particles generated re sulting in an increased lung deposition. Inhaler device perfor mance is routinely evaluated at inspiratory flow rates of 30 -120 L/min with 60L/min being considered as an optimal flow rate. It should be ensured that subjects with varying severity of asthma should be able to get enough pulmonary dose by generating sufficient inspiratory airflow rate [ 18]. In one study, the lung dose decreased from 22

PAGE 23

27.7% to 14.8% when the inspiratory flow rate decreased from 58 L/min to 36 L/min indicating the influence of flow rate on lung deposition [44]. The commonly used jet nebulizer is based on constant output design and is run by compressed air or oxygen. Supplemental air is dr awn across the top of the nebulizer. The drug may be present in the form of suspension or solu tion. Latest developments in the nebulizers have led to the introduction of devices based on electro-hydrodynamic principles. The underlying process utilizes electrostatic energy to create fine aerosols from the drug formulation. This leads to an increased pulmonary drug deposition. Am ong various types of nebulizers, the Mystic from Batelle provided pulmonary deposition as hi gh as 80%. Another nebulizer device called Respimat from Boehringer Ingelheim utilizes a high pressure micro-spray system of nozzles to slowly release a metered dose to the patient. Th is system also results in spraying a high concentration of respirable particles. With better devices, one can achieve greater pulmonary targeting by controlling different factors, the most important of which is PSD. In case of ICS which have considerable oral bioavailability, this may prove be neficial as smaller dose may be required with greater fraction of dose deposited in the lung. Apart from this, the side effects that one may experience from the high oropharyngeal deposition can be minimized w ith better pulmonary targeting. In case of newer corticosteroids with <1 % oral bioavailability, the dos e swallowed may not be of consequence as the systemic exposure is de pendent only on the lung dose [45]. The delivery devices are patent protected and these patents are often valid for longer periods of time than the patent for the drug ingredient. This makes it di fficult to have a device very similar to the innovator inhaler. 23

PAGE 24

Issues Involved in Bioequivalence Testing The reference or brand formulation is gene rally the original ma nufacturers product. Bioequivalence indicates similar rate and extent of drug absorption from a test and reference formulation at the site of ac tion and is aimed to ensure that two different products when administered in equal doses by same route a nd dosage form yield same pharmacodynamic (PD) effect [46]. There are standard ized methods for drugs administ ered orally and parenterally. However, this is not the case fo r bioequivalence of inhalers. Bi oequivalence of inhalers would require demonstrating similar rate and extent of drug absorption in th e biophase-lung. The main aim of the bioequivalence study for inhalation drugs is to ensure that the brand and the generic inhalers deliver similar amounts of drug, with similar regiona l deposition and similar lung residence times. The regional depo sition is of importance since we saw in the above discussion that asthma is a disease of the lower and upper airways. Any study designed to compare inhaled products should be able to discriminate between two the inhalers with re spect to these criteria. Needless to mention is the fact that the generi c inhaler should demonstrate a safety profile similar to the brand inhaler. It will be eviden t from the following chapters how difficult it is to evaluate bioequivalence of inhaled corticosteroids. 24

PAGE 25

Oral Bioavailability (F) % Protein Binding % Clearance L/hr Vd L Half Life hr Beclomethasone dipropionate 15 87 230 20 0.1 Beclomethasone monopropionate 26 NA 120# 424# 2.7 Budesonide 11 88 84 183 2.8 Ciclesonide <1 99 152 207 0.4 des Ciclesonide <1 99 396# 1190# 3.6-5.1 Flunisolide 7 80 58 96 1.6 Fluticasone propionate <1 90 69 318 14.4 Mometasone furoate <1 98 54 332 4.5 Triamcinolone acetonide 23 71 37 103 2.0 Table 2-1. Pharmacokinetic propert ies of inhaled corticosteroids NA Not Available, Vd Volume of Distribution, # apparent maximum approximation based on complete conversion of the parent compound into metabolite 25

PAGE 26

Figure 2-1. Activation of anti-inflammatory ge ne expression by corticosteroids (Barnes [30]) Figure 2-2. Fate of inhaled corticosteroid 26

PAGE 27

CHAPTER 3 POTENTIAL TOOLS Introduction Factors relevant for evaluating bioequivalence (BE) of inhaled corticosteroids (ICS) are pulmonary drug deposition, pulmonary residence time and the central to peripheral deposition (C/P) ratio in lung. Ahrens et al. have propos ed to test the BE of inhalation products by measuring clinically relevant pharmacodynamic (PD) endpoints such as FEV1, PEF and FEF in asthmatics. However, with corticosteroids, th e high variability and quantitative insensitivity of such PD endpoints might necessitate large asthmatic population to establish BE. The regulatory authorities from different countries have diffe rent views as to what constitutes sufficient conditions to demonstrate BE of ICS. Though there has been a change in the approach by these authorities, some regulatory agencies such as the US FDA are still in the process of formulating them. What needs to be remembered is that th e requirements should be ra tional and feasible. As no single approach suffices the need of time and cost effective technique to establish BE, a combination of in vivo and in vitro techniques may prove to be useful. This chapter reviews guidelines by regulatory authoriti es from different countries a nd discusses in detail options available. Potential tools A number of methods are availa ble and being developed to ev aluate the bioequivalence of inhalation drugs. The primary characteristic needed with any evaluation method is that the tests should be able to distinguish between different inhalers based on amount and rate of drug being available in the lung. These tools involve pharmacodynamic (PD) or clinical efficacy trials pharmacokinetic (PK) studies scintigraphy or imaging techniques in vitro formulation characterization 27

PAGE 28

This chapter reviews in detail various tools and discusses them in light of the latest developments. The PD or clinical efficacy studies aim to compare two inhalers to ensure that they have similar efficacy. The PK studies use th e plasma concentration time profile data to answer BE questions. One of the important in vitro tests include studies us ing cascade impactor to compare the respirable drug dose. Cascade imp actor tests can also be used to get a more detailed analysis of the partic le size distribution (PSD) of the aerosol generates by the inhaler. The PSD directly affects the regional drug deposition within the lung. The smaller the particle size, the deeper it will be able to travel within the lung airways that narrow as they branch into smaller and smaller airways. We will also disc uss the attempts to analyze the cascade impactor study data to answer certain questions involved in bioequivalence studies. Pharmacodynamic Studies The pharmacodynamic (PD) studies are in vivo clinical efficacy studies that measure and compare clinical endpoints after administering drug with innovato r and generic inhalers [46]. The PD endpoints such as forced ex piratory volume in one second (FEV1), peak expiratory flow rate (PEFR), forced vital capacity (FVC), and forced expiratory flow (FEF) are measured in asthmatics using spirometry. Subjective measuremen ts may be recorded by the patients using diary cards. In this diary card, the patients make a note of the events such as coughing, sleep disturbances due to asthma attack, chest ti ghtness score, how often rescue medication (bronchodilator) was used. There are other tests to detect the drug induced changes in nonspecific airway hyper responsiveness (AHR). AHR can be measured following the inhalation of direct bronchoconstrictors ( e.g. histamine or methacholine) or indirect agents such as adenosine-5'-monophosphate (AMP). Directly acting agents activate cholinergic receptors in bronchial smooth muscle and cause airway narrowing. Bronchoconstric tion by the indirect 28

PAGE 29

agents is a result of degranulation of mast cells which results in release of proinflammatory mediators [47-49]. This test, also known as PC20 or PD20 measures the concentration or dose of these bronchoconstrictors that br ing about a 20% drop in the FEV1. High reproducibility has been established with this test [50]. ICS have shown to attenuate the AHR provocated with these molecules. Thus, these endpoints have been used as clinical outcomes w ith ICS therapy [51, 52]. A large number of studies have used these e ndpoints to evaluate th e performance of anti asthmatic drugs in mild to moderate asthmatic subjects. The information obtained from such bronchoprovocation studies has sim ilar clinical relevance as th at obtained from bronchodilation studies. Airway hyperresponsiveness induced by AMP is regarded as a more reliable model for the evaluation of anti-asthmatic effects of ICS ra ther than the direct bronchoconstrictors [53]. Budesonide, fluticasone propionate and ciclesonide have been shown to improve AHR induced by AMP [54-56]. However, the duration of action of these drugs in this model has been rarely studied [56]. Also the changes in drug induced AHR induced by corticosteroids is modest [57]. Spirometry tests as well as other categorical PD endpoints such as sleep disturbance scores, wheezing scores, shortnes s of breath, and frequency of 2 agonists usage to maintain asthma stability are also evaluated when comp aring the ICS. Asthmatic subjects also show increased levels of exhaled nitric oxide (eNO) [57, 58]. Treatment with anti inflammatory agents like glucocorticoids resulted in d ecrease in the eNO levels [59]. It has been suggested to use NO levels as a biomarker to monitor the underlying inflammation in as thma symptoms [60]. However, the use of eNO to assess or modify as thma therapy with inhaled glucocorticoids has not resulted in any significant success [61, 62]. More information is needed on use of eNO in evaluating glucocorticoids to treat underlying inflammation in asthma. 29

PAGE 30

Any PD test employed to compare glucocorticoids (or any other drug) must be able to differentiate the responses between two clinically relevant ICS dose levels. In the absence of this property, the test may demonstrate that two inhalers to be equiva lent even if they delivered different doses. All of the above discussed PD endpoints fail to e fficiently differentiate between different dose levels of ICS. There exis ts a lack of well defined and immediate pharmacodynamic response with ICS. More than 34 weeks of dosing period is required to see any significant effects of ICS in improvement in asthma or maintaining the stability of an existing asthma condition. Such lengthy treatment duration makes the parallel study design a viable option. With para llel study designs, BE clinical trials between corticosteroid inhalers would necessitate study populations in excess of hundreds of patients with at least around 3-4 weeks of ICS dosing [63, 64]. The sensitivity an d variability of the selected pharmacodynamic end points toward the ICS play a deciding role in whether a dose -response can be shown or not [65]. In the past, a large multi-center trials have b een conducted for periods of about two to four weeks (some even up to twelve w eeks and greater) to elucidate the dose-response for ICS. These trials have used parallel study designs. Only a fe w studies were able to demonstrate statistically significant differences between di fferent dose levels [66]. Table 3-1 highlights some studies where absence of dose-response can be seen at various dose levels. In the study by Pearlmen and coworkers, 327 subjects with m oderate asthma were given eith er placebo, 50, 100 or 250 g of fluticasone propionate for a period of 12 weeks. FEV1 and FEF were the PD endpoints measure. The study could detect no differences between any of these dose levels [67]. In another study by Singh et al. more than 300 subjects received 400 and 800 g of budesonide over 12 weeks. There was no difference in the PEFR measured [68]. In another PD study, Peden et al. compared 50 30

PAGE 31

and 100 g of fluticasone propionate in more th an 400 subjects with ch ronic asthma. The study failed to differentiate the two dose levels with th e FEV1 and PEFR tests [69]. It can be seen in these studies summarized in table 4-2 that even up to 4 fold dose differences cannot be detected with the PD studies. This can be attributed to the low sensitivit y of the lung function tests to assess the differences between different dose levels with inhaled corticosteroids. Difference in individual responses to inhaled corticosteroids is illustrated as one of the reason of the flat doseresponse curve. Some patients respond to very small doses of ICS ach ieving maximum benefit whereas some of them are steroi d resistant. A study with a mix of steroid sensitive and steroid resistant population can confound the results of such studies. Use of steroid nave population is recommended in studies comparing inhaled co rticosteroids to avoi d such issues [33]. The PD studies represent dose response studi es where the reference and test drugs are given at different dose levels to human. At least one formulation is given at more than one dose level. This will help establish a dose response curve. The other formulation can then be compared against this curve. Hence, in essenc e the PD studies depict the human studies to estimate relative potency rather th an clinical efficacy studies [70] However, this approach has met with only limited success. The ratio of the do se response slope and th e variability of the responses, (variability/slope) determines the pow er to estimate the rela tive potency between two formulations. A small ratio will mean low res ponse variability and high dose response slope and will provide greater power to estimate relative po tency [71]. In case of ICS, the shallow doseresponse curve and high variability in responses makes it difficult to use such a PD approach. Crossover designs may help to reduce the variab ility in the responses seen in parallel study designs. Also, the use of steroid nave patien ts may overcome some limitations of steroid resistance [66, 71]. This may result in a smaller variability/slope ratio an d consequently greater 31

PAGE 32

power in estimating the potency ratio of inhalers as discussed above. The crossover study design gives rise to the issue of carryover from previous study arm. This is due to the fact that the effect of ICS in maintaining asthma stability conti nues for sufficient time post discontinuation of the ICS. Ahrens et al. proposed a cross over desi gn to study the clinically relevant pharmacodynamic responses to evaluate the BE of the generic inhalers. The authors initially employed the study design to compare the efficacy of inhaled -2 agonist inhalers [72-75]. These studies employed two dose levels of each of the generic and innovator inhalers being compared. The 2 agonist mediated inhibition of bronchospasm induced by methacholine or histamine was used as the pharmacodynamic endpoint. The Finney 2X2 bioassay statistical procedures were used to estimate the relative pote ncy of the generic inhaler relati ve to the innovato r inhaler [76]. This method estimated the relativ e number of actuations of the innovator (standard ) inhaler to yield approximately similar effect as one actuatio n of the generic (test) inhaler. The approach was utilized to demonstrate the BE of an al buterol generic inhaler to the innovator inhaler Ventolin [77]. The authors applied the same approach to estimate relative potency on glucocorticoids inhalers (hence indirectly eval uate BE of the inhale rs) Beclamethasone dipropionate (BDP) [78]. Ahrens et al. proposed a different approach to avoid this carry over effect. The study scheme is depict ed in figure 3-1. Brie fly, asthmatic subjects were screened and entered into the trial. There was a 5-14 day run in period to check compliance with data recording. During this period the s ubjects used their regular ICS. Af ter this the subjects were pre dosed for 4-7 days with 40 mg twice daily dose of prednisone. This was done to bring the patients to maximum possible corticosteroid response in spirometry tests. The FEV1 was checked and then the subjects started their treatment with either the innova tor or test (generic) 32

PAGE 33

inhaler for three weeks. At the end of three weeks, a histamine chal lenge was performed along with other spirometry tests. After this, the subject entered th e next study arm which started by pre-dosing with prednisone for 4-7 days. This wa s done to avoid the carry over effect in crossover design [71]. The outcomes were then analyz ed to determine the dose response curve. The authors suggested that the test and reference drugs be dosed for more than three weeks. The endpoint was maintenance of asthma stability ra ther than improvement in the asthma condition. The authors assert that using th eir proposed study design, a BE of glucocorticoids inhalers could be evaluated using fewer (30-50 subjects). A complete study evaluating innovator and generic inhaler using this approach is yet to be c onducted. Additionally, such a cross over design with multiple dose levels of each inhaler would lengthen the study period over several months. This could result in poor study compliance. There is lack of consensus with regards to optimal study design. There are issues with parallel and cross over study desi gns as we saw in the preced ing sections. Finding sufficient number of steroid nave patients wi th a predefined level of asthma (generally mild to moderate) is difficult. The PD approach t hough enticing and intuitiv e fails to establish BE of inhaled drugs in a time and cost-effective manner. After a more clear review of literature studies, it is obvious that such an emphasis by the re gulatory authorities is unfounded and unfeasible. Such PD studies take patients in excess of hundreds, lengthy trial periods and still fail to distinguish between dose levels of ICS. To summarize, the pharmacodynami c approach with ICS is difficult and has been attributed to their shallow dose-response relationship, poor selection of dose ranges, high variability of the responses, lack of well defined site of action for ICS in asthma as well as to the limitations of the study design employed [79, 80]. 33

PAGE 34

Pharmacokinetic (PK) Studies Pharmacokinetic studies involve measuring plasma drug concentrations after inhalation. Basic PK parameters such as area under the curve (AUC) and maximum plasma concentration (Cmax) are used to evalaute systemic drug expos ure. While AUC is a measure of the cumulative drug exposure to the body, Cmax is a measure of the rate of drug absorption into systemic circulation. In case of inhalation drugs, PK studies have been view ed to be of limited use in the past. This was due to the limitations of the analyt ical techniques to detect the low plasma drug levels that one encounters after drug delivery via inhalation. However, PK of most inhaled glucocorticoids can be assessed with assa y techniques like LC-M S/MS and tandem mass spectrometry. The ICS drug levels in plasma can now be measured in the low picogram levels [81-83]. Hence, plasma levels after inhalation do nt seem to be a limiting factor as reasoned out in the past. Absorption of the swallowed drug fraction may occur from the GI tract as discussed in chapter 2. Charcoal block is utili zed to prevent the oral absorption of the fraction of the drug that is deposited in the oropharyngeal region and may be swallowed [84, 85]. Essentially, one conducts a PK study with and without charcoal administration. The charcoal suspension adsorbs most of the ICS that enters the GI tract and prevents it from being absorbed in the systemic circulation. The newer ICS like fluticasone pr opionate, ciclesonide and mometasone have negligible (< 1%) oral bioavail ability and any plasma concentration would be reflective of only pulmonary deposition [86-89]. Hence, charcoal bloc k is unnecessary with such ICS. Plasma level for these drugs with negligible oral bioavailabil ity reliably reflects the amount of drug absorbed from the lung only. Thus, the plasma levels of ICS after inhalation are indicative of the amount deposited in the lung. Attempts have been made to compare the lung deposition of ICS based on plasma levels after inhalation. A study demonstr ated equivalent lung deposition after inhalation 34

PAGE 35

of single dose of budesonide from test and refere nce DPIs using the PK data [90]. Studies have compared inhalers by employing more than one tool such as PD study, PK data, imaging techniques and in vitro tests. In the ensuing review, we will see examples where PK data detected differences in drug exposure when PD ap proach failed to do so due to the shallow dose response relationship. Singh et al compared the equivalence of budesonide in healthy and asthmatics, children and adults using PK as well as PD measures [68]. The study demonstrated equivalence between two pMDIs (based on CFC and HFA propellants) using PK data. The study compared single and multiple doses with 3 dose levels as well as varying number of inhaler actuations. There was no difference in the plasma drug levels between the CFC and HFA pMDIs with single or multiple doses. Dose proportionality in AUCs was observed w ith all the dose levels selected (400 g 1600 g). Interestingly there were no statistica lly significant differences in pulmonary function endpoints, asthma symptoms, sleep disturbance or use of rescue medication between the two inhalers at two dose levels (400 g and 800 g). T hus, from the PK data (AUC) we see that there was different drug exposure between 400 and 800 g dose levels. Hence, the 400 g and 800 g dose levels were not bioequivale nt. If only a PD endpoint was to be considered, the inhalers would have been wrongly qualified as equivalent. Another study by Daley-Yates et al. compared two in-house inhalers, each delivering a combination of 50 g of salmeterol and 250 g of fluticasone propionate [91]. One inhaler was a multiple dose dry powder inhaler DISKUS while the other was a reservoir powder inhalation device (RPID). PD, PK and in vitro comparisons were made. The difference in the mean change from baseline morning peak expiratory flow rate (PEF) was used as the PD end point. The study was conducted for 12 weeks and 270 subjects co mpleted the study. There was no statistically 35

PAGE 36

significant difference observed in this efficacy endpoint. In vitro particle size distribution (PSD) comparison was made using 8 stage Anderson Cascade Impactor. The PSD profiles were comparable for the two inhalers. Interestingly, the PK da ta revealed that there was more than 2 fold systemic exposure to FP with RPID as compared to DISKUS. The estimated AUC ratio for FP exposure was 2.00 with 90% CI 1.56-2.55 This shows that the PK approach has greater sensitivity in detecting the differences between inhalers with regards to dose delivered to the lung. One of the biggest advantages of using the PK approach is that it can be performed in healthy volunteers, obv iating the necessity of asthmatics. With ICS like budesonide the in vivo PK in healthy volunteers is sim ilar to that in asth matics [92, 93]. There are some references which demonstrate lower plasma levels for drugs like fluticasone propionat e in asthmatics. This has been attributed to the high lipophilicity of fluti casone propionate and in such drug specific cases further studies might be necessary [92, 94] With ICS exhibiting slow absorption, other PK parameters such as mean residence time (MRT) and mean absorption time (MAT) could be estimated and compared. We hypothesize that differ ences in regional drug di stribution within the lung can also be estimated with PK studies for su ch drugs. PK studies have been consequential in explaining pulmonary drug deposition and unders tanding the therapeutic equivalence [95-97]. The safety profile of the drug is directly correlated to the drug exposure that can be estimated from the plasma levels. Hence, a PK study w ould be sufficient to estimate the drug exposure (and hence efficacy) and safety profile as well. The preceding discussion highlights the importance of PK studies to measure the pulmonary drug deposition. Such a PK approach can be used to estimate the pulmonary as well as the total body exposure. The number of days and subjects required for such an approach will 36

PAGE 37

be greatly reduced in comparison to a clinical efficacy trial. A PK study can be completed over few days in a cost effective manner. PK appro ach under appropriate conditions exhibits immense potential in the design of BE studies for inhalers. Scintigraphic Studies Drug deposition in the central and peripheral ai rways is of critical importance in asthma. The site of action of anti asthmatics is poorly defined. The inflammatory process in asthma occurs throughout the lung and hence the deposition of corticoste roids (anti-inflammatory) is essential throughout the bronchopulmonary regions [98]. There have been arguments that regional drug deposition cannot be appreciated with PK data. We hypothesize that this is true only for fast dissolving drugs. Hence, differen tial drug deposition in the central and peripheral regions within the lung for such fast dissolving drugs may be required to elucidate maximal response or efficacy. Scintigraphic studies for inha led medications can accomplish this task. The use of drug attached with a radioactive tracer has been traditionally employed to determine the spatial or regional drug distribut ion within the lung after inhalation [99]. It involves radio labeling the formulation with ra dionuclide such as technetium 99mTc or other gamma emitting isotope which is then inhaled by the subject. Occas ionally, inert carrier par ticles, such as Teflon particles have also been used instead of the drug in such depo sition studies. Two-dimensional planar gamma images are taken and compared after subtracting for background levels. The central, intermediate and peripheral pulmonary deposition is compared between the test and reference product. The total lung deposition can also be calculated from the scintigraphic images. This technique suffers from that fact that it is often not possible to radi olabel the drug molecules as such. Change in the radio-labe led particle behavior as compared to the original particle has also been argued. Validation of the in vitro radio-labeling experiments are needed to demonstrate that the aerodynamic PSD of ra dio-labeled product is simila r to original product [100, 101]. 37

PAGE 38

There also may be overlap of regions due to the two dimensional natu re of the image. Variability between different observers is also unavoidable. Recent developments in imaging techniques like single photon emission comput ed tomography (SPECT) and positron emission tomography (PET) have led to techniques in whic h three dimensional images can be obtained [102, 103]. Currently these imaging methods are under development to overcome some shortcomings. With increasing popularity, they ma y prove of critical importance in comparison of deposition profiles in the future. Correlation of Lung Deposition Studies An obvious question is whethe r there exists a correlation between lung deposition and efficacy or response of inhaled drugs. Lung deposition serves as a measure of local drug bioavailability in the lung. Hence, with drug delivery to the site of action, improved lung deposition should result in an increased efficacy or response [63, 101]. There are reviews of studies discussing the positive correlation betwee n lung deposition and clin ical effect [18, 104]. These studies are classified into two broad categories. One type of study is where an equivalent control of asthma with a lower dose of drug from a device can be directly related to the improved pulmonary drug delivery from that particular device. The other types of studies demonstrate direct correlation between the e fficacies of inhaled drug with its lung deposition for two or more treatment regimens. It must be emphasized that studies with concomitant assessment of lung deposition and pharmacodynamic effect provide str ongest evidence. However, such studies are scarce [101]. A clinical trial with two dos e levels, 400 g/day and 1600 g/day of budesonide from a pMDI and a pMDI attached with a spacer (Nebuha ler) in 35 asthmatics concluded that a given level of antiasthmatic response could be achie ved at half the dose when Nebuhaler was used. This was attributed to the higher pulmonary depos ition resulting from the efficient delivery from 38

PAGE 39

Nebuhaler [105-107]. With improved lung depositi on there was reduction in oropharyngeal drug deposition and consequently reduced occurren ce of oral candidiasis [107, 108]. Numerous studies have been published demonstrating the highe r efficiency (almost twice) in terms of lung deposition of drugs from a DPI (Turbuhaler) as compared to a pMDI. One such study with Turbuhaler showed that 400 g by pMDI and 200 g by Turbuhaler were equipotent, and similarly 100 g by pMDI and 50 g by Turbuhaler were equipotent. The study concluded that half the dose of salbutamol from a Turbuhaler produced a similar bron chodilator effect when compared with pMDI. The improved efficiency of drug delivery to the lung thus correlated to lung function improvement [109]. There exists a paucity of such studies correlating lung deposition and clinical efficacy with the ICS with most of them involving 2 agonists [18]. Another PK study demonstrated a direct co rrelation between lung deposition and the pharmacodynamic effect with terbutaline, a 2 agonist [110]. Along with the absolute values fo r lung deposition, the va riability associated with it is also of particular interest. The vari ability in lung deposition can be explained by the variability in throat deposition and vice versa [111]. The vari ability is a measure of the range of lung deposition of ICS that can be expected with dail y use in individuals. This information will be of great importance whilst assessing the treatment regimens [111]. The variability in dose deposition to the lung could be assessed by in vitro as well as in vivo experiments. A few factors to be mentioned are device performance, device handling, patient co-ordination of actuation and inhalation, inhalation flow, patient throat anatomy etc [112]. A numb er of studies have attempted to account for the absolute values of lung depos ition as well as the in ter-individual and intraindividual variability associat ed with the amount deposited [85, 112, 113]. Studies have also 39

PAGE 40

compared the lung deposition and variability associat ed with it using different inhalers like DPI and the pMDI. The aforementioned studies provide sufficient evidence to suggest that lung deposition data could be used as surrogate marker for clinical response data under appropriate considerations. Consequently, PK studies in conjunction with the in vitro or scintig raphy studies, could be a useful tool to esta blish the BE of inhaled drugs. This data could be used to demonstrate the effectiveness of inhaled drug products to the regulatory agencies [101]. The number of patients as well as duration of study will be greatly reduced by using such a surrogate marker [114]. Significant time and financial reso urces could thus be conserved during the drug development and approval process using this a pproach. This approach could substitute the necessity of the inconsequential clinical efficacy trials. Two way cross-over PK study over single days could be sufficient for BE of ICS as against 4 weeks or more of parallel or months of crossover study designs in large patient popul ation with poor or confounding outcome [101]. This approach could also be useful in special cas es like regulatory approvals after change in the manufacturing site [115]. The outcome of this approach in cases where the delivery characteristics from the inhaler product are dissimilar needs to be carefully monitored. In vitro Studies The preceding sections discussed various in vivo techniques. Partic le size distribution (PSD) is one in vitro measurement that yields information to compare two inhalers. Spray pattern and plume geometry are in vitro tests used to characterize th e performance of the valve and actuator. Plume geometry can be measured usin g laser diffraction and tim e of flight aerodynamic particle size analysis. Microsc opy is also conducted to study the particles and their aggregates. One can get information about parameters such as mass median aerodynamic diameter (MMAD), geometric standard deviation (GSD), fine particle mass (FPM) and fine particle 40

PAGE 41

fraction (FPF) from the PSD. All these parameters ar e used to describe the particle size of the aerosol generated by the inhaler. Cascade Im pactors (CI) and multistage liquid impingers (MSLI) are used to study the PSD. The FPF represen ts the fraction of the so called respirable fraction (representing bronchopulmonary depositio n) of dose and usually are particles or droplets with MMAD of 1-5 um [116, 117]. It has been demonstrated that lung deposition efficiency is correlated to the MMAD and GSD [118, 119]. Improved effi cacy of the inhalation therapy can be achieved by selective deposition of the particles or droplets generated by the inhaler. The Anderson eight stage cascade impactor is extensively used to assess PSD, MMAD and GSD and is the device of choice [120]. The US F DA does not consider it adequate to compare PSD of two inhalers in terms of MMAD, GSD or the FPF but encourages the individual plate comparisons of particles in al l size ranges in the cascade im pactor [121]. There exist some limitations in precision to compare particle ma sses on a plate by plate basis. This has been explained and related to the ma nufacturing tolerances in the plate pore cut-off points [122]. Consequently the plate by plate comparison is highly variable and im practical. Plate grouping analysis and use of model fitting software (to f it bimodal or more complex PSD) is suggested to overcome some of these limitations [122]. The sel ection of induction port influences the outcome of PSD measured by cascade imp actor [123, 124]. Using the same cascade impactor to measure the test and reference inhaler (c ross-over design) helps to reduce the variability of PSD. As of now, the in vitro tests are considered to be more of quality control significance by the US FDA and are not sufficient to demonstrate BE of OINDP by themselves. They are important for product characterization and performance. This may be attributed to the fact that the in vivo 41

PAGE 42

inhalation process may be more complex. A simple metal port inductor used in cascade impactor may not be a good replication. This is we ll explained by the st udy discussed above. In the above mentioned study by Daley-Yates et al two in-house inhalers, each delivering a combination of 50 g of salmeterol and 250 g of fluticasone propionate [91] were compared. PD, PK and in vitro comparisons were made. The difference in the mean change from baseline morning peak expiratory flow rate (PEF) was used as the PD end point. The study was conducted for 12 weeks and 270 subjects completed the st udy. There was no statistically significant difference observed in this efficacy endpoint. The in vitro particle size distribution (PSD) was compared using a 8 stage Anderson Cascade Impactor. Interestingly, the PK data revealed that there was more than a 2 fold systemic exposure to FP with RPID as compared to DISKUS. The PK based parameter such as AUC ratio for FP exposure was 2.00 with 90% CI 1.56-2.55. Thus similar PSD profiles as determined by cascade impact or may not lead to similar lung deposition in vivo Summary Most of the spirometry tests as well as other clinical end points (P D markers) have high variability. The issue of steroid sensitivity wi thin study subjects further complicates the study design. With the newer drugs that have negligible oral bioavailability, the plasma levels are a reliable measure of the dose deposited in lung. Advances in the analytic al techniques make it easy to accurately estimate the low plasma drug levels. For ICS with significant oral bioavailability, alternative techniques like char coal block, can help differentiate between systemic drug levels resulting from lung or GI tract. This aids the reliable estimation of pulmonary drug deposition after inhalation. The in vitro comparison such as similar PSD may not necessarily translate in vivo similarity. Imaging techniques are still under development phase. In such situations, PK studies are better suited as compared to the PD studies to detect 42

PAGE 43

differences between ICS. Not only is this approach cost-effective but also saves time. It is better suited to detect differences between two inhale rs than the PD method. Consequently, the PK approach provides a tool that can effectively be used to evaluate whether a generic inhaler will provide similar efficacy and safety as that of the brand inhaler before such switch is made. 43

PAGE 44

Table 3-1. Summary of few studies with di fferent dose levels Study Drug & Doses Design and Duration Population/ Sample size (total) Spirometry PD endpoints measured Doseresponse observed Pearlmen et al FP 50,100 & 250 ug Placebo controlled, parallel, 12 week Moderate asthma/ 327 subjects FEV1 FEF (am/pm) No Singh et al BUD 400 & 800 ug HFA and CFC pMDIs Parallel, 12 week Adult and adolescent asthma patients/ 321 subjects PEFR (am) No Peden et al FP 50 & 100 ug by Diskus and Diskhaler Placebo controlled, parallel, 12 week Children (4-11 yrs) with chronic asthma/ 437 subjects FEV1 PEFR (am) No 44

PAGE 45

Figure 3-1. Ahrens design of cross over clinical study for BE of ICS 45

PAGE 46

CHAPTER 4 REGULATORY RECOMMENDATIONS There still exists a lack of consensus between industry, academ ia and regulatory authorities over criteria to establish bioequi valence (BE) of inhalation dr ug products. Regulatory agencies are in the process of finalizi ng a definitive guidance document about inhaler bioequivalence. Currently, two organizations, the European Medicines Agency (EMEA) and Health Canada have published guidance documents for establishing the bioequivalence of inhalation drugs. In this chapter we discuss these guidan ce documents. The US FDA viewpoi nt will also be discussed. Health Canada The Canadian Health Ministry released a draft guidance fo r a generic inhaled corticosteroid product in 2007 [125]. This document resulted from a long ongoing series of discussions through a series of symposia over more than a decade. The Canadian Thoracic Society with the support of Health Canada held the following three symposia that comprised international experts Seattle, Washington, U.S.A.May 1995 Toronto, ON, CanadaDecember 1995 Toronto, ON, CanadaMay 2000 The discussions from the first two sym posia were published in 1998 [126]. The publication [126] favored a pharmacodynamic approach over the direct pharmacokinetic measurement. The efficacy study represents a relative potency study that could be accomplished with clinical studies conducted in a number of ways. This pub lication elaborately discussed the various aspects including pharmaceutical factors that should be included in the guidance document in such comparative studies. The disc ussions from the third symposia were published in 2003 [127]. It summarized some of the possi ble approaches with their advantages and drawbacks. The various clinical study designs discussed (Table 4-1) were essentially human 46

PAGE 47

relative potency tests to determine the relative potency of the ICS. Th e two main study design approaches used either a controlled or uncontroll ed asthma model. There was a consensus in all the three symposia that there exists a need for a validated me thod to compare relative potency of ICS. Following the three symposia, a guidance docum ent was published by Health Canada that states the requirements for BE studies with IC S [125]. As per the document, two trials are necessary for establishing bioequivalence for the generic inhalers. The firs t is a well controlled, double blinded, randomize study with three parallel arms: reference product, generic inhaler and placebo. This resembles a Phase III efficacy study. The use of steroid nave mild asthmatic subjects with the lowest possible dose is recommended. Minimum of three weeks of trial duration is recommended in a parallel study de sign. The absence of clinically significant differences in the comparison of an inflammatory marker and pre-br onchodilator FEV1 is required. The guidance document clearly states th e criteria required when comparing the two inhalers: A difference in the mean sputum eosinophil coun t (expressed as a percentage of the total count) of at least 50% between the active trea tment (pre-treatment minus post-treatment, for both test and reference) and the place bo treatment (pre-treatment minus post treatment) will be considered clinically significant. A difference in the mean FEV1 (expressed as percentage of predicted) of at least 10% between the active treatments (post-treatment minus pre-treatment, for both test and reference) and the placebo treatment (post-treatment minus pre-treatment) will be considered clinically significant. The second trial required is a single dose PK study at upper dose limit to measure systemic exposure as a surrogate for long term systemic ef fects. In the absence of reliable analytic techniques in case of low blood levels, the draf t guidance allows for evaluation of systemic exposure by measuring the hypothalamus-pituitary-a xis (HPA) function. This entails measuring 47

PAGE 48

the serum cortisol over 24 hours with either a single or multiple dose study design. The summary of these requirements from the draft guidance are enlisted in figure 4-2. The draft guidance document lists other details that are require d to be submitted for generic drug approval. However, it may be difficult to evaluate the BE of ICS with a single dose efficacy study. With difficulties in differentiating between different dose levels of ICS based on PD endpoints, it seems a daunting task to detect differences be tween inhalers based on a single dose level. The difficulties with using the PD approach have be en reviewed in the previous chapter 3. European Medicines Agency The committee on medicinal products for human use of the European Medicines Agency (EMEA) has published a draft guideline on bioequivalence requirements for orally inhaled products in 2007 and 2009 [128, 129]. The suggested approach from this guideline can be summarized as in figure 4-3. The guideline requires only an in vitro equivalence study to be substantiated with regards to drug, excipients and container pe rformance criteria. Such an in vitro test can be conducted with us e of multistage impactor. The following creiteria for the in vitro tests needs to be fulfilled. The criteria incl ude that the generic product contains the same active substance in an identical dosage form. This means that a pMDI should be compared to a pMDI and a DPI to a DPI. There should be similar instructions for the us e of the two inhalers. The differences in the excipients (if any) shoul d not influence the product performance, aerosol particle behavior or the inhalation behavior of the patient. Similarities between device performances such as resistance to air flow, dose delivered are required as well. The guidance requires that either the in dividual stages are directly compared or the stages are divided into at least four groups. Comparison should be made for th e stages that represent fi ne particles, as well as the upper stages of the impactor. In the event of failure to establish the in vitro equivalence, an in vivo study is required to substant iate the equivalence of ICS. 48

PAGE 49

Deposition studies comparing the extent and pattern of pulmonary deposition could be done by imaging or PK studies. The imaging study c ould be done either by using two or three dimension scintigraphy. The regional pulmonary deposition can be compared by measuring the radioactivity in the different segments of the l ung. A PK study is still re quired as a measure of systemic safety. A PK study may be used al one to assess the pulmonary deposition. The study design for such a PK study should be able to exclude drug absorption from the GI tract. In the event of this approach failing too, a Phase III PD efficacy trial will be required to investigate the therapeutic equivalence of the inhaled drug products as per th ese guidelines. The clinical endpoint should be a pulmonary function measur e and preferably FEV1 although other endpoints such as expired nitric oxide (NO) PC20, PD20 and sputum eosinophils. The choice of efficacy endpoint needs to be justified based on its sensitivity to detect differences between adjacent doses. A minimum of eight weeks of dosing is recommended with a pa rallel study design as concerns are expressed over th e carry over effect in cross ove r study design. There is emphasis on the selection of homogenous study population t op minimize the variability in response which would thereby increase the power in detecting dose response relationship and differences in formulations if any. The EMEA published an updated guideline do cument recently 2009 [128]. The most important update in the latest document was in the BE test requirements of ICS for pediatric population. The document quotes that the pulmonary deposition studies(PK and imaging) are not appropriate in children. Pharmacokinetic studies as a surrogate for effi cacy only imply efficacy, they increase the burden on the child and have insufficient advantages over pharmacodynamic and/or clinical studies in 49

PAGE 50

the assessment of therapeutic equivalence in ch ildren to warrant their use. Imaging studies in children are also not appropriate. With children aged 6 years and older, the clinical efficacy endpoint for the pharmacodynamic study should be a pulmonary func tion measure such as FEV1. PEF may be used with appropriate ju stification. In child ren, the systemic safety can be compared with a PK test if sufficient information about effect of reference drug on child HPA axis is available. In the absence of this information a pharmacodynamic safe ty test is required. This includes measuring the effect of drug on the HPA axis and on the lowe r leg bone growth rate as a surrogate marker for growth. The EMEA presents a rather over-optimistic approach to evaluate BE of inhaled drug products. In vitro particle size distribution comparison by itself may be inadequate. There are instances where similar PSD in vitro may not necessarily transl ate into similar drug amounts in vivo The study by Daley-Yates et al was discussed above where such a situation arises [91]. The two dimensional imaging techniques have limite d advantage to differentiate between regional depositions. The three dimensional imaging tec hniques are still in development stage and posses cost constraints. In our view, one needs a PK approach supported by in vitro tests for example using cascade impactor to answer specific questions that arise with regards to slowly and fast dissolving glucocorticoids is better suited as compared to the rather ambitious concept of establishing BE of inhalation drugs on in vitro tests only. United Kingdom A workshop report from the British Associat ion for Lung Research (BALR) has been published in 1994 [120]. The overall recommendations are outlined in figure 4-1. This workshop was aimed by the British Agency to formulate guidance for BE of inhalation drugs. With the current EMEA guidance in effect, the UK documen t is obsolete. A combination of clinical 50

PAGE 51

efficacy studies as well as in vitro tests was recommended though not specified explicitly to demonstrate bioequivalence of inhaled medicat ions. The BALR workshop recommended that pharmacodynamic studies be conducted to compare the clinical efficacy of the inhalers in a controlled population. With ICS, studies for at least one m onth of drug dosing are recommended. The recommended PD endpoints include peak expiratory flow rate (PEFR am and pm), reduction in response to bronchoprovocation (PC20), improvement in the frequency of bronchodilator use, sleep disturbances and fewer exacerbations. Th e workshop considered PK studies of limited value and inessential in determin ing inhaler bioequivalence due to low plasma levels of ICS. Spacer devices and their combinations complicate the assessment and PD studies with and without spacer are recommended for their comparison. United States Food and Drug Administration Currently, there is no guidance from the Un ited States, Food and Drug Administration (US FDA) for BE of inhalation drugs. This is unlik e for the nasal products where the US FDA has published a draft guidance in 2003 [130]. The US FDA is actively seeking to establish BE a guidance for inhaled products. There have been ongoing discussions by the Orally Inhaled and Nasal Drug Products (OINDP) subcommittee about th is issue. The discussions encompass issues with BE of nasal sprays, inhale rs that include bronchodilators, corticosteroids and as well as combination products. The subcommittee convened a meeting in 2000 (Rockville, Maryland) with the aim of devising guidelines for BE st udies [131]. However, no consensus was reached about what appropriate tests were required. The committee agreed that the overall approach should be to establish BE of inhaled drug products by demonstrating equivalence based on in vitro tests, clinical efficacy study and systemic exposure safety study. The various PD study designs discus sed included the study 51

PAGE 52

design proposed by Ahrens et al [71]. The design has been discusse d in detail in the previous chapter. The study design was used by FDA to ev aluate the BE of the inhaled bronchodilator, albuterol. The relative bioavailability is estimate d in terms of the dose of the generic inhaler required to produce the same effect as that of the reference inhaler. The FDA has funded the follow up study with ICS using Ahrens approach at University of Iowa. The committee also expressed interest in exploring th e use of exhaled nitric oxide as a surrogate marker for treatment with ICS. The FDA has also funded a study to be conducted at National Jewish Medical and Research Center in Denver, Colorado. The International Pharmaceutical Aerosol Consortium on Regulation and Science (IPACRS) and the scientists of th e Inhalation Technology Focus Group (ITFG) of the American Association of Pharmaceutical Scientists (AAPS) have been collaborating to address specific issues for the OINDP. The Product Quality Res earch Institutes (PQRI) Working Group (WG) has published the results (2007) of the chi-square test statistic used to compare the aerodynamic particle size distribution (A PSD) from cascade impactor data [132-134]. Summarizing their findings in the final paper, the WG pointed ou t the limitations of using the chi-square test statistic as well as its combination with the supplemental population based equivalence (PBE) in comparing the APSD cascade impactor profiles [134]. Regulatory Approach Comparisons The regulatory authorities in different c ountries have made recommendations to demonstrate inhaler bioequivalence [ 120, 121, 126, 135]. A detailed study of the recommendations from these regulatory agencies highlights the lack of consensus and varied approaches taken by regulatory authorities in di fferent countries. While the Canadian Regulatory requirements find the in vitro tests inconsequential, the EMEA is ready to authorize generic inhalers that pass the in vitro tests alone. The UK, US FDA and H ealth Canada stress the need 52

PAGE 53

for a clinical efficacy study and tend to minimize the relevance of pharmacokinetic studies. This emphasis on conducting a clinical efficacy study sounds intuitive but upon further reasoning seems unfeasible. These reasons with examples ha ve been discussed above in detail. This has been mainly attributed to the poor dose-respons e curves for the inhale d corticosteroids. The required duration for such efficacy trial differs fr om country to country ranging from three to eight weeks minimum. The clinical endpoints that need to be compared also differ between various regulatory agencies. A lthough, all allow the crossover st udy design, a parallel design is preferred to prevent carry over effects. The selection of low dose and steroid nave patients is something which is not easily met. The require ment of single dose study comparison as per Health Canada seems irrational when one consider s the flat dose response curve with ICS. The two dose level requirement with EMEA guidelines may not be required if the inhalers pass in vitro tests. There are cases when in vitro similarity does not necessarily translate into in vivo equivalence [91]. There still lacks a definitive un iform guideline from these agencies that could be used by the pharmaceutical manufacturers to get their products approved. All in all, the relevance of conducting a PK study is assigned to evaluate the safety or systemic exposure of the ICS. The requirements range from being too dari ng as with EMEA to no guidance as seen with the US FDA. Summary There are ongoing efforts everywhere to provid e a robust and reliable method to establish BE of inhaled products. These efforts are active at all levels including the academia, industry as well as the regulatory authorities. The regulatory authorities recomm end clinical efficacy studies, safety studies in all the drug classes for inhale d drugs. There exists some contention over what constitutes a sound clinical study de sign in terms of clinical end poi nts, patient selection criteria, study duration, clinical and statistical significance of the elucidated re sponse for BE of inhaled 53

PAGE 54

products. However, finding such a universal time and cost-effective study design to establish BE of inhaled drugs has so far resembled the search of Holy Grail with every attempt turning futile. The drug classbronchodilators or ICS differ in their mechanism of action and need to be targeted in different regions in the lung. It might be possible to perform clinical efficacy trial with the bronchodilators, although even these dr ugs easily tend to reach the dose-response curve plateau. There exist inherent difficulties in clinical trials to compare the ICS due to the flat doseresponse curves. Most of the spirometry tests as well as other clinical end points (PD markers) have high variability. The issue of steroid sensitivity within st udy subjects further complicates the study design. With the newer drugs that have neg ligible oral bioavailabil ity, the plasma levels are a measure of the dose deposited in lung. The lu ng deposition studies can be used as a bridge between in vitro and clinical studies in drug development. Advances in the an alytical techniques and using alternative techniques like charcoal block, make it eas y to accurately estimate the plasma profiles which reflect lung deposition of the drug. This aids the reliable estimation of pulmonary drug deposition after inhalation. In vitro PSD comparisons as well as lung deposition have a great potential to be used as a surrogate marker for clin ical response studies. In such situations, PK studies have an upper edge over th e PD studies. Not only is this approach costeffective but also saves time. It is better suited to detect differences between two inhalers than the PD method. Consequently, PK approach provides a tool that can effectively be used to evaluate whether a generic inhaler will provide si milar efficacy and safety as that of the brand inhaler before such switch is made. Conclusion It can be concluded that each of the PK, PD, scintigraphy or in vitro approaches has its own advantages and disadvantages. Any approach ca nnot be used as standa lone to establish the BE of inhaled medications. A combination of these methods can be employed and their 54

PAGE 55

importance is directly related to the drug as we ll as the drug product. While it may seem possible to perform the clinical study with some drugs like 2 agonists, this approach is fraught with difficulties and has been inefficient in case of IC S. This has been proved by the numerous failed efficacy trials that can be studied from the pa st. We hypothesize that th e study design with PK method in conjunction with appropriate in vitro tests may be one of the solutions to unravel the BE conundrum with inhaled drug products. Any appr oach should be devised with the intent to provide quality drug products to the patient in a cost-effective way and timely manner. The PK approach poses the potential to fulfill this requirement. Hypothesis We hypothesize that combination of in vitro and PK studies are sufficient to evaluate the BE of ICS. Such an approach requires few healthy subjects and provides a time and cost effective solution for BE studies of such inhalation drugs. All the three questions of how much drug is available, where does it get deposited and how fast the drug is absorbed can be answered successfully with this approach. 55

PAGE 56

Table 4-1. Clinical models to determine relative potency of ICS Clinical efficacy model Description Advantages Disadvantages Controlled Asthma Allergen provocation Measure airway responselate asthmatic response (LAR) due to allergen provocation Possible use at lower dose level (give references) Expertise needed to perform the provocation tests Shallow doseresponse Exercise/ methacholine/ AMP provocation Airway hyperresponsiveness to bronchial provocation Bronchial provocation by AMP more sensitive -Results not reproducible -Dose and duration of treatment critical Uncontrolled Asthma Natural -subjects with moderate to severe asthma subjected to treatment -subjects with mild asthma stabilized with high prednisone dose and treated to maintain asthma stability (Ahrens) -simple and relatively easy -Crossover design possible to reduce variability large sample size due to shallow dose response Shallow dose response and lengthy trial periods Steroid reduction Subjects with stable asthma on ICS have steroid dose reduction by 50% at definite intervals until asthma is not under control -model good for comparing different ICS 50% dose reduction occurs long duration confounding by natural exacerbation possible 56

PAGE 57

Figure 4-1. Recommendations for Bioequivalence of Inha led Medications by BALR 57

PAGE 58

Figure 4-2. Health Canada Draft Guidance Requirements for ICS Bioequivalence 58

PAGE 59

Figure 4-3. EMEA guideline scheme for bioequivalence 59

PAGE 60

CHAPTER 5 HOW MUCH DRUG IS AVAILABLE? Introduction: The ICS are intended to deliver the drugs in the lung where they exhibit their efficacy before being absorbed into the systemic circulation where they may exert their side effects. It would be very easy to demonstrate BE if one could obtain lung fluid samples (eg. bronchial lavage) and measure the drug concentrations in the lung. Another approach would be to demonstrate that both inhalers have similar effi cacy and safety when given at equal doses, which is the ultimate aim of demonstrating BE. The PD approach has difficu lties as discussed in chapter 3 which make it an inefficient tool to evaluate BE. The shallow dose response curve with ICS is one of the main reasons for the problems with the PD approach. We hypothesize that the pharmacokinetic (PK) approach with suitable in vitro technique can be used to demonstrate the bioequivalence of ICS. The PK approach involves using the plasma concen trations as well some information from the in vitro tests to answer the bioequivalence question. The questions that one needs to answer to demonstrate bioequivalence are how much drug enters the lung, where is it deposited with in the lung and how fast is it absorbed. The presence of mucociliary cleara nce as well as systemic absorp tion from lung necessitates studying how long the drug stays in the lung before being absorbed. In other words the residence time of the drug from the two inhalers al so needs to be compared. The question of how much goes into the lung arises because the dose delivered by th e inhaler in one actuation is not the dose deposited in the lung. Some terms that are generally used while de scribing dose from inhalers are discussed below. Nominal dose is the dose delive red by the inhaler in on e actuation as per the manufacturers claim. The ex-actuator dose is the dose which actually le aves the actuator in a single puff. There may be some drug loss in the spacer (if used). Most of the drug which enters 60

PAGE 61

the mouth is impacted on the throat (oral cavit y and oropharynx) and is swallowed (drug enters the GI tract). Only a small fraction of inhaled drug enters into the l ung. This fraction of drug which is the effective dose deposit ed in lung is termed as respirab le dose and can be expressed as a percentage of the nominal dose. This value is highly variable and generally ranges from 5-40% of the nominal dose. A very small fraction of this inhaled drug will be exhaled as well. The inhaler type, particle size genera ted from the inhaler, inspiratory flow rate, lung function, throat geometry, inhalation technique, use of spacer, inhale r priming are some factors that affect this fraction reaching the lung [112]. All the above factors related to the drug, inhaler and individual explain why the final amount reaching the lung c ould be very low at times as well as highly variable. We will discuss in sufficient detail in the following chapter about mucociliary transport process that exists in the upper or the central airways. This process can clear some solid drug particles out of the lung before they are dissolved in lung fl uids and undergo absorption into systemic circulation. This is a critical phenomenon for slowly dissolving drugs. Latest developments in inhaler technology (eg. dry powder inhalers) have successfully attempted to increase the amount of drug reaching the lung. At the same time these inhalers have also decreased the effort in terms of coordination re quired on part of the patient during inhalation. The plasma levels of ICS reflect the drug de posited and absorbed from the lung. This is due to the negligible (<1%) oral bioavailability of most of the IC S [86-89]. For drugs that display significant oral bioavailability one could use the charcoal block technique to estimate the drug absorption from the GI tract after being swa llowed [84, 85]. We hypothe size that the plasma levels obtained by a PK study after inhalation co uld be used to successfully answer all the questions pertaining to establ ishing the BE of inhalation dr ug products. We use modeling and simulation technique to test our hypothesis. This would be a better way to test the feasibility of 61

PAGE 62

conducting a PK study by simulating it first in silico to get better understanding of the outcomes from such a study. The modeling and simulations were done with reference to fluticasone propionate. In this chapter, only AUC is used to answer this specific que stion of how much drug is available to the lung. Methods The data to develop a PK model following administration of ICS was obtained from a previously published study comparing the sing le dose and steady state pharmacokinetics and pharmacodynamics of inhaled FP and BUD in healthy volunteers [136]. The study was conducted in accordance with the revised Declaration of Helins ki and in compliance with good clinical practice guidel ines. 14 healthy male volunteers co mpleted this double-blind, doubledummy, randomized, placebo-contro lled, 5-way crossover study. The subjects had a mean age of 26.4 years (range, 22-32), average weight of 72.7 kg (range, 62-85 kg), and average height of 179.2 cm (range, 172-187 cm). The five treatments consisted of 200 g FP and 500 g FP (both doses delivered via the Diskus inhaler, 400 g BUD and 1000 g BUD (both delivered via the Pulmicort Turbohaler, AstraZeneca), and place bo delivered via Diskus and Turbohaler. The subjects were administered single doses of drugs at 8.00 AM on day 1 and then twice daily from days 2-5. Blood samples were collected frequently on day 1 and day 5. On the other days, days 2-4 only the trough samples were collected i.e. samples collected at 8.00 AM and 8.00 PM. The samples were analysed for drug, cortisol and ly mphocytes. The details of this study can be found elsewhere [137]. Only the FP data for first 24 hours was used to develop the model. Population Pharmacokinetic (POPPK) Model: The population pharmacokinetic model was built using a step wise model building approach. The model was developed using the nonlinear mixed effects modeling software NONMEM VI (Globomax LLC, Hanover MD). The data was plotted to see any outliers as well 62

PAGE 63

as check for any other inconsistencies. Model fits with one and two compartment body models with first order absorption were evaluated. The individual fits, populations fits, minimum objective function value, residual plots were used to determine the best fit model. The interindividual variability (also known as inter-subje ct variability or between subject variability [BSV]) was modeled as exponentia l error model as equation 5-1 Pij = i exp( i) (5-1) where Pij is the i th parameter for the j th subject. i is the i th population mean parameter and i is the random intersubjec t variability with mean 0 and variance 2 The residual error (als o known as intra-subject variability or within subj ect variability [WSV]) was modeled as proportiona l error model as equation 5-2 Yij = Ypij*(1+ ij) (5-2) In the above equation, Yij and Ypij are the ith individuals jth ob served and model predicted concentration respectively. ij is the random residual error with mean 0 and variance 2. First order conditional estimation (FOCE) was used for parameter estimation. Individual parameter estimates were estimated post hoc The final model was bootstrapped using Wings for NONMEM developed by Dr Nick Ho lford to perform the internal validation of the model. The criterion to include a parameter (theta or eta) in the model was greater than 6.61 points reduction in the minimum objective function value within the hierarchal model. The likelihood ratio follows a 2 distribution and hence with df=1, and P<0.01 level of significance a drop of 6.61 points would make the added parameter signifi cant to be retained in the model. Simulations The above described POPPK model was used to simulate plasma profiles of FP following various scenarios of drug deposition in the lung. The differe nt scenarios tested were combinations of different amount of drug bei ng deposited in the lung as well as differing 63

PAGE 64

variabilities associated with this amount. The pulmonary dose lie s anywhere between 5-40 % of the nominal dose (in some cases it maybe up to 60%). However, this amount is also accompanied by huge variability of around 15-45% (in some cases it could be even higher) [111]. We simulated various scenarios by assigning differe nt amount of dose deposited by two inhalers representing the innovator (brand or reference) and generic inhale r. The plasma profiles were generated from these simulations. The area under th e curve (AUC) for every subject is calculated using the trapezoidal rule from these plasma pr ofiles. The AUC is a measure of cumulative drug exposure to the subject. The AUCs are log transf ormed and then the AUC ratios are calculated. The 90% confidence interval (CI) is then calculated and compar ed as done in traditional BE studies [138, 139]. The AUCs are said to be equi valent if the 90% CI of the ratio of the log transformed AUCs is between 0.8-1.25 (or 80-125% when expressed as percentage). There exists variability for all the parameters such as drug cl earance, apparent volume of distribution, amount deposited as well as drug absorpti on. We incorporated all the para meter variabilities as listed in Table 5-1 while testing the different scenarios with this model. To ensure that we have enough sampling for these variabilities, we simulated 200 runs for every scenario and then summarized the BE results for these 200 runs. The process was repeated for every scenario. The entire modeling and simulation process is depicted stepwise with a flow chart in fig 5-1. The trial was simulated using NONMEM while the huge amount of data generated from these simulations was analyzed by using the statistical software R. Results A two compartment body model was used to de scribe the plasma concentrations of FP after inhalation. Literature references have demons trated that FP exhibits flip-flop kinetics [140]. This phenomenon is observed when the absorption is the rate limiting step in drug disposition. The prolonged absorption of FP is also of intere st to study the residence time of FP and other 64

PAGE 65

ICS which exhibit similar absorption profiles [141]. The flip-flop behavior of FP was incorporated into the model and fitted the data well. The residual plots and population and individual fits were observed to assess the goodne ss of fits. It can be seen from the plots in figures 5-2 and 5-3 that there is no significant bias in the data The predicted population as well as the predicted individual con centration time plots fit the observed data very well. This model was further used to perform simulations under various scenarios. It must be remembered that the model assumed that the plasma levels reflecte d drug absorbed from lung only. This is a valid assumption since FP has less than 1% oral bioavailability [142, 143]. Consequently any drug swallowed during or post inhalation (due to muco ciliary clearance) will not be reflected in plasma. The spaghetti plots of simulated plasma prof iles are presented in figure 5-4. The results from the simulations are present in Table 5-2. Le ts discuss first scenario where both inhalers deliver equal amount of drug (500 g) in the lung w ith 30% variability. It can be seen from run # 1 that with 24 subjects, 60% of the trials co uld establish equivalence. The equivalence is established if the 90% CI of th e ratios of the AUCs of the gene ric and brand inhaler are within 80-125%. The considerable variabilities on various parameters necessitated more subjects in the study. Hence the same scenario was simulated w ith greater number of subjects. From run numbers 2-5 it can be seen that with increased subjects, the percentage of trials that could demonstrate equivalence improved and gave gr eater confidence in th e outcome. With 40-50 subjects, equivalence could be established in more than 90% of trials. PK trials with scenarios that had greater variabilities on dose deposited in the lung were also simulated. As is evident from runs 6-9 with equal dose delivered to lung equivalence could be esta blished in most cases with sufficient number of subjects in the study. 65

PAGE 66

Another scenario where the gene ric inhaler delivered a different dose as compared to the brand inhaler to the lung was si mulated. In this case, the gene ric inhaler delivered lower dose than the reference inhaler. Such a scenario could be attributed to a lot of factors such as larger aerosol particles, poor performance by the inhale r device. In this partic ular case, the generic inhaler was simulated to deliver 20% lower dose as compared to the reference inhaler. Runs 1011 demonstrate that with 30-40 subjects majority of the trials fail to establish equivalence. In cases where there were actual significant differences between the doses deposited in the lung, increasing the number of study subjects would still result in failure to es tablish bioequivalence. In fact the difference would be easier to capture with an increased number of study subjects. This will result in greater confidence in the trial outcome. Conclusion From the above simulations we prove our hypothe sis that AUC is a sensitive parameter to answer the question as to how mu ch drug is available to the lung. This parameter can thus be used to compare the inhalers with respect to the amount delivered to the lung. There exists a theoretical exception to this for slowly dissolvi ng drugs such as fluticasone propionate. One may argue that the regional deposit ion pattern may be adjusted wi th a lower or higher dose from generic inhaler and yet have comparable AUCs. Although highly unlikely, theroretically it may be possible to have such a scenario. This represents a special case and is disc ussed in detail in the next chapter and appropriate measures that c ould be employed in such scenario are proposed. Discussion ICS are the best way to deliver the corticos teroids to the site of action. Establishing bioequivalence with inhaled cort icosteroids requires demonstrat ing equivalent dose deposition, comparable regional deposition and similar resi dence times with the innovator and generic inhaler. Each part of the above question repr esents a different chal lenge. The pharmacokinetic 66

PAGE 67

approach seems to present a potential solution to th is issue. In this chap ter we discussed the PK approach to answer the first part about how much drug enters the lung. The simulations demonstrated that this answer could be successf ully answered by using th e plasma concentration time profiles of the drug from a PK study. Such a pharmacokinetic approach utilizes the plasma levels which are not as invasive or uncomfortable as obtaining lung fluid samples. The assumption that the plasma levels reflects only drug deposited in lung is true for newer ICS like ciclesonide and mometasone as they have less than 1% oral bioavailability [144, 145]. ICS like beclomethasone and budesonide exhibit some degree of oral bioavailability (10-30%) [146, 147]. Concomitant oral administration of activated charcoal to prevent GI absorption is one approach to account for the swallowed fraction of dose. [85]. In this method, a charcoal suspensi on is administered to the subject. The drug that enters the GI tract is then adsorbed into char coal and does not enter th e systemic circulation. Another approach involves estimation of lung deposition after inhalati on by correcting for an assumed oral availability. By simulating different scenarios, it can be seen that the question of how much drug goes into the lung can be succe ssfully answered by this approach. The high variability may warrant a few mo re subjects as long as the am ount of drug from two inhalers does not vary significantly. In cas es with significant dose differ ences with high variability, the model identified correctly the bioequivalence to fa il. The above observation is intuitively correct that if the same amount of drug is dispensed from two inhalers and only variability is high then one needs higher number of subjects to correc tly identify whether differences exist. The approach is similar to the traditional BE studies with oral dosage forms and can be completed over few days including the washout periods between crossover arms. There may be minor adjustments that one might make depending on the ICS drug characteristics (eg. charcoal 67

PAGE 68

block). A big advantage with such approach is the use of healthy volunteers to demonstrate the BE unlike the pharmacodynamic approach where one cannot use healthy volunteers. Also the number of subjects required for such an approach is very small (30~50). The PD approach would require 100s and 100s of patients (depending on the study design), to be dosed over extended periods and yet fail to achieve any significant co nclusion. We already discussed this in the 2nd chapter. Research and studies have shown that ther e may be differences in the amount of drug reaching the lung in healthy volunteers and asthma tics. It seems obvious that in asthmatics the pulmonary obstruction will cause such a differenc e. Due to narrowed airways resulting from pulmonary congestion, most of the inhaled drug will be deposited in the central airways and not reach the peripheral region within the lung. Another reason suggested for such an observation is the drug characteristics. For example, such a di fference in regional depo sition is more marked and been reported for slowly dissolving drugs such as fluticasone propionate. The high lipophilicity of FP results in slow absorption fr om the lung. The mucociliary action which is predominantly active in the central airways clea rs the drug out of the lung which is either expectorated or swallowed. This results in an ov erall reduction in drug dos e available in the lung. In the next chapter we discuss this issue and s ee how PK approach can be used to answer this question. A theoretically probabl e scenario for slowly dissolving drugs where AUCs may be inconclusive to answer the BE questions is also discussed in that chapter. Finally we will also discuss how PK approach could be used to answ er the question of comparing the residence time of drug in the lung. Any approach should be devise d with intent to provid e quality drugs to the patients in a cost effective and timely manner. 68

PAGE 69

Table 5-1. Parameters and th eir associated variability Paramater Variability CL 10% Vd 10% ka 30% Flung 30% Residual 20% 69

PAGE 70

Table 5-2. Results from simulations run # # of subjects % Variability on dose % Dose-Generic % BE Trials successful % Dose-Brand 1 24 30 500 60 500 2 30 30 500 78 500 3 40 30 500 90 500 4 50 30 500 95 500 5 75 30 500 100 500 6 50 35 500 87 500 7 75 35 500 99 500 8 50 40 500 74 500 9 75 40 500 94 500 10 30 30 400 4 500 11 40 30 400 1 500 70

PAGE 71

Figure 5-1. Flow chart for mode ling and simulation of trials 71

PAGE 72

Figure 5-2. Goodness of fit plots. The open circles are the observed values and the lines are model fits 72

PAGE 73

Figure 5-3. Individual and popul ation fits from the model 73

PAGE 74

Figure 5-4. Spaghetti plots of plasma concentrations from simulations 74

PAGE 75

CHAPTER 6 WHERE DOES THE DRUG DEPOSIT? Introduction The questions to be answered to demonstrate the BE of ICS are how much drug enters lung, where does it get deposited, how long doe s it stay there before being absorbed systemically. The previous chapter discussed how pharmacokinetic (PK) studies can assess how much drug is available to the lung. We also hypothe sized that the PK approach can be used to answer the remaining questions. We discuss in th is chapter the use of PK approach to answer where does the drug get deposited within the l ung central to periphe ral (C/P) lung deposition ratio. Post inhalation, the ICS is deposited in th e lung where it exhibits its efficacy. Eventually, the inhaled corticosteroids will either be absorbed into systemic circulation or will be cleared by mucociliary transport process [148, 149]. Attempts to capture and model such phenomenon are available in literature [10, 150]. A slow absorption process (conse quently greater residence time) leads to prolonged exposure and may translate into better efficacy. Prior to its activity at the receptor level, there may be additional proce sses occurring such as drug release from the formulation and dissolution in the lung fluids pr ior absorption [151] It has been known that there exists difference in the absorption rates of drugs from different regions in the lung. The absorption rate is higher in the alveolar region (AL) or the peri pheral region as compared to the tracheo-bronchial (TB) region or the central region [5, 152]. Some other processes such as drug metabolism may also occur in the lung [153]. The drug may also be cleared by mucociliary transport from the lung after being deposited. Thes e processes contribute collectively towards the duration of activity of ICS maki ng it a complex phenomenon [8, 10, 151]. 75

PAGE 76

Absorption and Muco-ciliary Clearance The mucociliary clearance and drug absorption are the two processes which predominantly affect the drug presence within the lung. These two processes th ereby characterize how long the drug stays in the lung before being cleared fr om the lung or absorbed into the systemic circulation. The absorption of IC S occurs after the inhaled part icles have undergone dissolution in the lung fluids. The mucociliar y clearance is very slow in th e peripheral region. This is well explained by the anatomy of the lung airways. In the central airways, the tracheal region along with the bronchial airways are ciliated and heavily lined by mucus producing goblet cells. The ciliary structures bathe and beat rapidly in a blan ket of mucus layer. The epithelia in this region are covered by mucus, lipids, glycoproteins, inorganic salts and water [154]. Airway mucus humidifies the inhaled air. It also entraps particulate matter, bacteria, viru ses, gaseous particles, and aerosols. The entrapped material is expector ated by the ciliary acti on thereby providing the lung a defense mechanism against foreign particul ate matter. The TB regi ons have considerable mucociliary clearance that can clear the particles in few hours. However, such clearance mechanism may be affected by clinical conditions such as asthma, bronchitis and cystic fibrosis [155, 156] While the drug is being cleared by mucociliary mechanism, drug absorption into systemic circulation also occurs simultaneously. The rate of absorption in the upper or central airways is small as compared to that in the lo wer or peripheral regions The ciliated and mucus producing airway structures star t disappearing as the bronchioles branch down further into terminal bronchioles and alveolar regions that are non-ciliated [4]. Consequently, the inhaled particles may be retained in the alveolar region for long periods of time (in days). At the same time the absorption from the alveolar regions is very high. The absorption rate from the peripheral compartment is twice as high as that from the central region [5, 6]. Hence, the inhaled drug particles (which are soluble in nature) will be absorbed be fore any significant clearance 76

PAGE 77

through mucociliary mechanism occurs in these peripheral regions of the lung. The rates of dissolution, absorption and muco-c iliary clearance thus govern the amount of drug in the lung. This is well explained with figure 4-1 that has been adapted from Edsbacker et al [7]. It can be seen that some particles in the ciliated airway s in the central region of the lung dissolve in the lung fluids (mucus layer) and are absorbed, while some particles are cleared by the ciliated cells before they are absorbed. Such a pheno menon is absent in the peripheral region. The mucociliary clearance rates as well as absorption rates differ depending on the lung region [8]. The mucociliary process may also be affected (slower) in subjects with lung diseases [9]. It has been rightly mentioned that determ ination of mucociliary clearance is difficult to quantify [10] and depends upon natu re of particles depositedsol uble or insoluble particles. Currently prescribed inhalation drugs can be consid ered to behave as solu ble particles. This can be construed such that the inhaled corticosteroids act at the receptor level at intra or extracellular locations in the lung, or may also be retained in interstitial fluid before being ultimately absorbed into systemic circulation. At the same time so me drug may be cleared by mucociliary mechanism [11]. The ciliary processes are predominantly activ e in the tracheo-bronchial region or the central region of the lung. Studies in the past have atte mpted to estimate the mucociliary clearance rates using radio-labeled particles such as liposomes, ferrous oxide part icles, carbon particles, Teflon particles as well as drug molecule s. These studies have found that th e tracer particles have a half life ranging from 0.5 hours to 24 hours depending on whether they are deposited centrally or peripherally within the lu ng [12-16]. It may take even days in some cases for the drug deposited in deep areas of the lung to be cleared. Howeve r, since the ICS are soluble and absorbable drugs, the drug deposited in the peripheral area of the lung will be absorbed before being cleared over such a long period of time. Hence only rapid muco-ciliary clearance which occurs from the 77

PAGE 78

central region of the lung woul d predominantly affect the amount of drug available for systemic absorption. Models of varying complexity have been us ed to study such drug deposition profiles. For simplicity of modeling, the lung could be consider ed to be made up of two regions, the central and the peripheral regions. The deposition patter n depends upon airway caliber, particle size distribution of aerosol generated and the lung function. The larger particles tend to be deposited in the upper airways while the sma ller particles can escap e to deeper or peripheral regions within the lung. The suboptimal lung function in asthma tic population leads to more proximal drug deposition. The narrowed airways cause more of th e drug to be deposited in the central airways than in peripheral airways. This has been explained graphically by Edsbacker et al in figure 6-1 [7]. Consequently the C/P ratio, which is the ratio of drug deposited cen trally Vs peripherally, could vary in healthy and asthmatics. Ideally one would want to deliver uniform amount of drug throughout the airways since asthma is said to be disease of upper and lowe r airways. With more central deposition in asthmatics there is more dr ug available for mucociliary clearance than in healthy volunteers for the same given lung dose. Added to this if the drug was slowly dissolving drug such as fluticasone propionate then th ere could be noticeable drug removal by such mucociliary processes. Hence less amount of drug actually is available systemically in asthmatics which could reflect as lower AUCs. Such a difference in AUCs between healthy and asthmatics has been reported in literature for slowly dissolving drugs like fluticasone [157-159]. We hypothesize that pharmacokinetic approach can be used to study the differences in regional deposition of drugs as well as the residence time of the drug in lung. As mentioned above, studies have also show n that there are differences in the drug distribution within the lu ng in healthy and asthmatic subjects This is attributed to reduced 78

PAGE 79

airway caliber in asthmatics. At the same time, the mucociliary clearance is decreased in the asthmatic population. The mucus viscosity increa ses and the ciliary processes are inhibited thereby lowering the drug clearance [155, 156]. Other factors such as the particle size distribution, device, nature of aerosol generated and individu al inhalation technique also influence the regional drug distribution. The regiona l distribution of drugs is measured in terms of the C/P ratio or the central to peripheral ratio. The C/P rati o is calculated by dividing the amount of drug deposited in the central region by the amount of drug deposited in the peripheral region. Several studies have used radio-labeling techniques to determine the C/P ratio using various inhalers in healthy and asthmatic subjects with differe nt drug molecules. These ratios range from 0.7 to 2 (or more) depending on the clin ical condition of the subject, the device used, particle size distribution and method used in the study [12, 160-162]. One would want comparable deposition in the central and peripheral regions of the lung with preferably slightly higher deposition in peripheral airw ays. Such a deposition pattern w ould give a C/P ratio of 1 or less. We hypothesized that the PK approach can be us ed to detect differences in the regional distribution of slowly disso lving ICS within the lung. Us ing population pharmacokinetic modeling techniques we test this hypothesis by employing PK trial simulations as demonstrated in previous chapter. Methods A population PK model was developed to describe the time profile of inhaled drug with the lung being divided into two dis tinct regions central and peripheral. Data from 30 asthmatic subjects was used to build the model. The study details can be found elsewhere [163]. The primary purpose was to build a model that coul d be used for simulation purposes. Our model incorporates the drug clearance from TB regi on with no clearance occurring from peripheral 79

PAGE 80

regions. The absorption rates in the central and pe ripheral regions were co nstrained to have two fold differences for reasons mentioned above. The population model was developed and this m odel was used to simulate different trial scenarios by varying parameters such as mucociliary clearance rates, actual dose deposited, variability on dose deposited, and C/P ratios. Details about the di fferent scenarios are mentioned later. Every trial was simulated 200 times to get enough sampling for the given parameter variabilities. The AUC and Cmax ratios were estimated for each simulation. The success or failure of every single trial was determined based on the AUC ratios. The population model was developed using NONMEM VI. NONMEM was also used for simulations of PK trials. The data obtained from simulations was processe d using the statistical software R. Simulations To study the effect of different regional drug deposition patterns as well as changes in the mucociliary clearance values, we simulated trials by keeping one parameter such as mucociliary clearance constant at a time and then varying the regional deposition patt erns and vice versa. Every trial was simulated 200 times to allow enough sampling for the given variability on all parameters. The results were also compared by simulating 500 times to se e if that resulted in different outcome. It was observed that having 200 runs was a good number with negligible changes in the outcome with 500 runs except fo r increased computing and post processing times. The BE success or failure is calculated for ev ery trial and then the number of runs that demonstrate BE are reported as a percen tage of total number of trial runs. The mucociliary clearance rates are discu ssed in the preceding section. Mucociliary clearance values of Kmuc of 0.5 /hr and 0.2 /hr were used for the simulations. With each of these mucociliary clearance values various depositi on scenarios were simulated. These scenarios varied from more peripheral deposition to more central deposition. The changes in either 80

PAGE 81

direction were in increments of 50 percent. When a majority of trials for a situation failed or passed, we checked this effect by increasing the number of subjects. The starting number of subjects for most of the cases was 30. This number was increased to 40 and eventually 50. In some cases, trials were simulated with 75 subj ects to check how the results changed with additional subjects. Table 6-1 mentions the variabili ty associated with various parameters in the model. Results The compartmental model is depicted in figure 6-2. There is negligible oral bioavailability for FP (< 1%), hence the plasma levels reflect the drug absorbed from the lung. Hence, the GI compartment was not included. Drug is availabl e for inhalation from the central and the peripheral regions of the lung. Part of the drug deposited in th e central region is cleared by mucociliary mechanisms while the remaining amount is absorbed into systemic circulation. Once absorbed, FP follows a two compartment model w ith rapid distribution in to peripheral tissues. Ultimately the systemically available drug is cleared from the central compartment. These systemic levels are used to generate the plasma concentration time profiles for the reference and the generic inhalers. Initially, the Kmuc of 0.5/hr and 30 subjects were used for the simulations. The C/P ratio was 0.82 which meant that about 45% of the dr ug entering the lung was deposited centrally while the remaining 55% of the drug was deposite d in the peripheral regions of the lung. The C/P ratios for the generic inhaler were then altered to give either a more central or more peripheral deposition as compared to brand inhaler. Th e dose delivered by the tw o inhalers was kept constant. The regional deposition from the bra nd inhaler was also alte red to test different scenarios. As we saw in the previous chapter th at the high variabilities necessitated study with 81

PAGE 82

more than 30 subjects, we also simulated the current study with 30 and 50 subjects. The results from these simulations are presented in table 6-2 Few scenarios from these simulations are disc ussed. It can be seen from table 6-2, run # 1and 2, the brand inhaler deposited 45% drug in the central and 55% drug in the peripheral part of the lung. In the same run we have a generic inhaler that has similar deposition pattern. With 30 subjects 82% of trials could establish BE and with 50 subjects 98% trials could establish BE successfully. In run # 3 and 4 we have a generic inhaler that has predominant drug deposition in the central region of the lung. In such scenario, majority of the trials fail to establish BE. Intuitively this makes sense, since for the same total lung dose a greater portion is in the central part of the lung. Hence, the amount being clear ed by mucociliary mechanism is greater as compared to brand inhaler and less drug is actual ly available for the lung. Hence inhalers should fail to demonstrate BE. Similarly when we ha ve a generic inhaler that deposits more drug peripherally (run # 11 and 12), there is less drug available that can be cleared by mucociliary process as compared to the reference inhaler. In su ch cases we see that majority of the trials fail to establish BE. Similar observations were made with other C/P ratios with the brand inhaler. This confirms that for slowly dissolving dr ugs, AUC can detect differences in regional distribution pattern. The same process was then re peated with the mucociliary clearance rate of 0.2/ hr. The results from those simulations are presented in table 6-3. Similar findings are observed with this mucociliary clearance rate. However, the sensitivity of differences in regional deposition decreases at lower mucociliary clearance rates. Greater than two fold differences in the regional deposition can be picked up from the AUCs. Special Scenario From the above simulation results we see that the AUCs are affected by the dose available and the regional deposition as well. A situati on may arise where the dose deposited and the 82

PAGE 83

regional deposition may be altered simultaneously. Such a scenario is theoretically possible but is highly unlikely practically. We simulated such scenarios and dem onstrated that inhalers with lower respirable dose and more peripheral depos ition have similar or comparable AUCs to inhalers with more respirable dose and more central depositi on. The difference in dose is adjusted by differences in regional drug depositio n. The respirable dose is the dose that reaches the lung postinhalation. The simulations for such special scenario are summarized in table 6-4. Run # 41 describes a generic inhaler that has more peripheral deposition (70%) as compared to the brand inhaler with only 50% peripheral depositi on. Consequently with the same respirable dose, dose ratio of 1, the AUC with generic inhaler wi ll be higher as seen (AUC gene ric/brand ratio -1.25). This is due to the fact that the brand inhaler has more drug deposited centrally th at can be cleared and hence not available. We see that in run # 42, the dose ratio is 0.8 with the generic inhaler delivering lower dose with more pe ripheral deposition as compared to the brand inhaler. This demonstrates that the doses could be adjusted wi th different regional drug distribution to give similar AUCs. There are two solutions to this scenario. One way is to conduct a PK study in asthmatics. Since the asthmatics have predomin ant central deposition, the two inhalers when compared in asthmatics could help identify these differences. We simulated various trials to see what C/P ratios would give similar AUCs with different doses in healthy volunteers and asthmatics. The assumption for the simulation was that the asthmatics have predominant central deposition. We see from run # 48 that the generi c inhaler would need to almost revert the deposition pattern to have comparable AUCs as brand inhaler in asthmatics. This seems highly unlikely that for the same generic inhaler the deposition patterns would be totally opposite. In 83

PAGE 84

fact even in most cases with greater deposition in upper airway s AUCs would not be similar as seen from runs 45 through 51 Another way to overcome this scenario would be to do an in vitro test such as using cascade impactor and compare th e respirable dose. This will let us know whether the dose available is different and one may not need to do the study in asthmatics. For slowly dissolving drugs, only the comparison of respirable dose should suffice. This may not be the case with fast dissolving drugs such as budesonide. For fast dissolving drugs the regional deposition may still need to be answered by a more detailed comparison of cascade impactor data. The primary reason being that the drug is absorbed fast be fore significant amount of drug is cleared through mucociliary process. However, most of the gl ucocorticoids including the new ones are slowly dissolving. Conclusion The PK approach could be used effectively to answer the question of where does the drug deposit in the lung, especially fo r the slowly dissolving drugs. In some scenarios where the dose and deposition pattern might be altered simultane ously appropriate tests could be conducted to tackle those situations. However such scenario though theoretically possi ble is highly unlikely practically. For fast dissolving drugs one needs to do a more de tailed analysis of cascade impactor data. This is to compare the particle size distribution (PSD) of aerosol generated by the brand and generic inhaler. This is because we know that the PSD governs the regional deposition of the drug within the lung. Such an approach requires not more than 30-50 subjects. Regional drug distribution has been one of the main con cerns about employing PK approach to evaluate the BE of inhalation drugs. PK data successfully addresses this concern for slowly dissolving drugs. 84

PAGE 85

85 Discussion The regional distribution for ICS is of impor tance since asthma involves inflammation of upper as well as lower airways. The aim with th ese drugs is not only to deliver drug to the lung but also achieve deposition thr oughout the lung. If the generic i nhaler can demonstrate that it delivers comparable drug dose with similar regi onal deposition and that the drug stays in the lung for same period of time, one can confid ently assume that th e generic inhaler will demonstrate similar efficacy as the reference inha ler. Evaluation of differences in the regional deposition seems difficult by using the PK data. Ho wever, the differences in absorption rates, the slowly dissolving nature of most glucocorticoids and the pr esence of mucociliary clearance in upper airways can be effectively used to answer this question. Appropriate tests and studies can be used to supplement such PK study. The aim of such supplemental tests is to avoid remote possibilities of approving substandard product. The se nsitivity of the PK a pproach is better than the clinical efficacy trial. The PK approach provides an effective tool that can better compare the brand and generic inhalers to reach a correct deci sion with regards to thei r bioequivalence in a timely and cost effective manner.

PAGE 86

Table 6-1. Variability on various parameters in the simulated POPPK model Parameter Variability Cl 10% Vd 10% Ka 30% C/P ratio 40% Kmuc 20% Flung 30% Residual 18% 86

PAGE 87

Table 6-2. Simulation results with Kmuc~0.5/hr Same dose available through reference and generic inhaler. Slowly di ssolving drugs such as FP Generic inhaler Brand inhaler run # # of subjects Central Peripheral C/P ratio % BE Trials successful C/P ratio Central Peripheral 1 30 45 55 0.82 82 0.82 45 55 2 50 45 55 0.82 98 0.82 45 55 3 30 63 37 1.70 6 0.82 45 55 4 50 63 37 1.70 5 0.82 45 55 5 30 70 30 2.33 0 0.82 45 55 6 30 75 25 3.00 6 1.50 60 40 7 40 75 25 3.00 5 1.50 60 40 8 50 75 25 3.00 4 1.50 60 40 9 30 30 70 0.43 28 0.82 45 55 10 50 30 70 0.43 35 0.82 45 55 11 30 22 78 0.28 6 0.82 45 55 12 50 22 78 0.28 3 0.82 45 55 13 30 33 67 0.49 19 1.00 50 50 14 50 33 67 0.49 20 1.00 50 50 15 30 25 75 0.33 3 1.00 50 50 87

PAGE 88

Table 6-3. Simulation results with Kmuc~0.2/hr Same dose available through reference and generic inhaler. Slowly di ssolving drugs such as FP Generic inhaler Brand inhaler run # # of subjects Central Peripheral C/P ratio % BE Trials successful C/P ratio Central Peripheral 16 30 45 55 0.82 86 0.82 45 55 17 50 45 55 0.82 99 0.82 45 55 18 30 63 37 1.70 34 0.82 45 55 19 50 63 37 1.70 50 0.82 45 55 20 30 70 30 2.33 9 0.82 45 55 21 30 75 25 3.00 38 1.50 60 40 22 50 75 25 3.00 59 1.50 60 40 23 30 30 70 0.43 52 0.82 45 55 24 50 30 70 0.43 70 0.82 45 55 25 30 22 78 0.28 27 0.82 45 55 26 50 22 78 0.28 32 0.82 45 55 27 30 33 67 0.49 47 1.00 50 50 28 50 33 67 0.49 65 1.00 50 50 29 30 25 75 0.33 3 1.00 50 50 88

PAGE 89

Table 6-4. Simulation results for special scen ario. Change in the drug dose and regional deposition pattern. Slowly di ssolving drugs such as FP Generic Inhaler Brand Inhaler run # Central Peri % BE trials success C/P Ratio Central Peri Dose Ratio AUC ratio C/P Ratio 41 30 70 0.43 7 1.00 50 50 1.0 1.25 42 30 70 0.43 84 1.00 50 50 0.8 0.99 43 80 20 4.00 0 1.00 50 50 1.0 0.64 44 90 10 9.00 0 1.00 50 50 1.0 0.52 45 60 40 1.50 0 9.00 90 10 0.8 1.38 46 65 35 1.86 1 9.00 90 10 0.8 1.28 47 70 30 2.33 22 9.00 90 10 0.8 1.19 48 75 25 3.00 80 9.00 90 10 0.8 1.10 49 80 20 4.00 76 9.00 90 10 0.8 1.00 50 87 13 6.69 16 9.00 90 10 0.8 0.85 51 90 10 9.00 4 9.00 90 10 0.8 0.80 89

PAGE 90

Figure 6-1. Different centra l and peripheral deposition in healthy and asthmatics 90

PAGE 91

Figure 6-2. Compartmental model for inhaled fluticasone propionate used for simulation 91

PAGE 92

CHAPTER 7 HOW LONG DOES DRUG STAY IN LUNG? Introduction In the previous chapters, pharm acokinetic (PK) approach was used to evaluate how much drug is available to the lungs and where does th e drug deposit in the lung. The final question is how long does the drug stay in the lung before bei ng absorbed into the systemic circulation. This is important since the goal of as thma inhalers is local lung delivery of the glucocorticoids. With all other factors being held cons tant, the longer the drug stays in the lung the greater will be the efficacy. However, this would also mean that th e drug might be cleared out of the lung by the mucociliary action before interacting with th e receptors [11]. Once the drug molecule has dissolved and interacted with receptors, it will be absorbed into systemic circulation. Drug dissolution into lung fluids from deposited aerosol is a rate limiting step before systemic absorption occurs. This process of drug dissolution may be a characteristic of the drug. It may also be influenced by a slow release formula tion. The dissolution of drug molecule from the formulation blend in the generated aerosol is di rectly influenced by the formulation excipients. With the mucociliary mechanism in lung, part of the drug may be cleared before being dissolved in the lung fluid. Consequently, there may be less drug and hence change in efficacy as compared to the reference product. Hence, one may argue that even formulation differences may affect the drug profile in terms of the reside nce times in the lung. While one may measure the amount of drug available in lung by using a metric such as the AUC as we saw in previous chapters, one also needs to show that the rate at which drug is available and the duration for which it stays in lung is same between products to establish BE. This relates not only to the efficacy but also the drug safet y. Ultimately, one needs to evaluate if there are differences between the times that the drug st ays in the lungs when delivered by brand or generic inhalers. 92

PAGE 93

There are certain PK parameters that characterize the duration of time a drug resides in the body. The mean residence time (MRT) is such a PK parameter that describes the average amount of time a drug molecule resides in the body. MRT is thus influenced by the drug absorption and clearance. Mean absorption time (MAT) is anothe r PK parameter that could be used to answer residence times for drug in the body. It describes the average time required for the drug to be absorbed into systemic circulation. To be more specific, MAT refers to the time a drug molecule spends at the site of absorption before being absorbed into kinetic space. MAT thus gives an estimate about the time the drug stays in the lung before being systemically available. For example, one may measure MAT after oral or inhalation route. MA T by any route can be calculated as difference of MRT of a drug following administration by that route and intravenous route as explained in equation 7-1. MATinhal = MRT inhal MRTi.v. (7-1) In the traditional bioequivalen ce (BE) studies for oral ge nerics, one can find maximum plasma concentration Cmax being compared among formulations. This can be understood well when one reads BE definition which requires equal rate and extent of absorption. The Cmax term reflects the rate of absorption. This term could be used with inhalation drugs as well since for drugs with same dose, similar Cmax would mean similar rate of drug being available for absorption from lung site. With inhalers that have similar respirable dose, similar Cmax could demonstrate that the residence times of the drug in the lung are same. We wanted to test which of these terms would be better suited to evaluate the lung residence time of the drug in the lung. In other words, we wanted to test the sensitivity of these various parameters to answer the final question in evaluating the inhaler BE. 93

PAGE 94

Methods and Simulations The model used for in chapter 5 was used in this case as well. The details for model building can be found in the met hods section of chapter 5. With the inhalation route the dose is delivered into the lung from where it is systemi cally absorbed. The abso rption process involves the dissolution step before the drug interacts with the receptor and gets systemically absorbed. Hence the absorption rate constant is a hybrid co nstant that reflects the processes of dissolution as well as absorption. Once abso rbed, the drug is explained by a two compartment body model. The model was simulated to generate plasma concentration time profiles in various trial scenarios. The simulated scenarios involved brand and generic inhalers wi th different absorption rates. As discussed above, the different absorp tion rates incorporate diffe rences in dissolution rates as well. The population PK software NON MEM was used to run the trial simulations. Sufficient variabilities, as one might expect in a clinical setting were added to the model parameters. Table 5-1 lists the variabilities used for the simulations. Every scenario simulation was run 200 times to have sufficient sampling for these variabilities. The plasma profiles were processed using R software. Statistical moment an alysis was used to calculate PK MRT. These MRT calculations can be explai ned by equations 7-2 to 7-4 AUC = (Conc) dt (7-2) t 0AUMC = (Conc time) dt (7-3) t 0MRT = AUMC/AUC (7-4) We see from equation 7-1 that to calculate MAT we need MRT data after intravenous drug administration as well. MAT is then the diffe rence between the MRTs. Hence the model had three arms where the subjects were given drug by inhalation route, generic and brand, and by IV 94

PAGE 95

bolus injection in the third arm. The MAT and MRT were compared similarly as the AUC and Cmax parameters. Essentially, MAT and MRT were log transformed and similar comparison made based on the limits of 0.8-1.25 Results The model that was setup can be explained by figure 5-1. We see that there is an intravenous dose arm in addition to the inhaled do se arms. The results from these simulations are presented in table 7-1. The respirable dose available is same for both these inhalers. They have the same variability on all parameters as well. The only parameter that was altered was the absorption rate constant. As discussed above, changes in absorption rates may be seen due to changes in dissolution profile of the formulati on as well. We see that for run # 52, with 30 subjects and similar absorption rates, 76 % of the trials have equivalent AUCs. For the same scenario, more than 72% of th e trials have equivalent Cmax values. 100% of the trials were equivalent with respect to MAT and MRT comparisons. These results are expected since the inhalers have identical properties with respect to dose and absorption rates. A bigger study size with 36 and 50 subjects was simulated in runs 53 and 54 respectively by keeping everything else same as in run # 52. We now have a highe r percentage of trials having AUCs and Cmax equivalent. Subsequent study scenar ios were simulated with 36 subj ects as we determined from run #2 that they had sufficient power. However for this particular question we were more interested in evaluating the Cmax, MAT and MRT and hence we now focus only on these parameters in this chapter. In run # 55, the absorption rate ra tio is 2 and we see that none of the trials had equivalent Cmax or MAT. Around 40% of trials still s how similar MRT. In runs 56and 57, MRT is equivalent in all the trials despite different ab sorption rates for the two inhalers. The Cmax and MAT on the other end have most of the trials failing. The sensitivity of MRT to detect 95

PAGE 96

differences was poor. Small changes in absorption rates do not affect MRT significantly as MRT is a parameter that is influenced by the drug ab sorption and drug clearance. Consequently with drug clearance remaining unchanged, MRT is not a ffected significantly with small changes in MRT. MAT was also not reliable in cases wher e absorption rates differed. These parameters do not give us any better comparis on than Cmax which could sti ll be used successfully as it correctly failed to show any equivalence when the absorption rates differed. Estimation of MAT requires an intravenous drug dosing arm in the study. This results in increased time and cost of study with no added benefit to the outcome. Conclusion The residence time of the drug in the lung (and the rate of absorption) could be compared using the traditional PK parameter Cmax. Parameters such as MAT and MRT are not reliable in evaluating inhalers to answer th e question of how long does the drug stay in lung (how soon is it available). Cmax can be used successfully to answer that question. Around 30-50 subjects are sufficient to compare this parameter. One needs no intravenous study to estimate the Cmax values. With all questions being answered, we propose a BE study design for inhaled glucocorticoids in the following chapter. Discussion The residence time of drug in the lung is infl uential factor in the drug efficacy. The longer a drug stays in the lung, the greater will be the e fficacy. A slower release will lead to better lung targeting. One needs to remember that the mucocili ary transport that exists in the upper airways could affect the amount of drug available if the release is very slow by removing the drug before it dissolves and interacts with th e receptors. The drug lipophilicity is also of consequence since post release from formulation, the drug needs to dissolve in the lung fluids. For immediate release systems, the formulation blend will affect the drug release. Lactose is a commonly used 96

PAGE 97

97 excipient. However, differences in the manuf acturing process could affect the drug release profile. The results from simulations in this chapter suggest that Cmax is a better parameter to compare inhalers for the residence times. Cmax also is important in terms of drug safety. This brings up an entire discussi on of the limits needed for Cmax. While Cmax may be of critical importance for certain drugs, the equivalence limits could be widened for drugs that have broad therapeutic index. Such wider limits could be employed for inhaled corticosteroids. This is especially true in case of highly variable drug products.

PAGE 98

Table 7-1. Simulation results for inhalers with different absorption prof iles. Slowly dissolving drugs such as FP Parameters with their % equivalent trials run # # of subjects Ka-Generic Ka-Brand Ka ratio AUC Cmax MAT MRT 52 30 0.21 0.21 1.0 76 72 100 100 53 36 0.21 0.21 1.0 87 84 100 100 54 50 0.21 0.21 1.0 96 96 100 100 55 36 0.21 0.11 2.0 47 0 0 40 56 36 0.21 0.15 1.4 82 1 9 100 57 36 0.21 0.28 0.8 84 2 22 100 58 36 0.21 0.42 0.5 83 0 0 14 59 50 0.11 0.21 0.5 44 0 0 47 60 30 0.31 0.21 1.5 86 1 0 100 61 50 0.42 0.21 2.0 87 0 0 7 98

PAGE 99

Figure 7-1. Compartmental model scheme for in haled and intravenous fluticasone propionate used for simulation 99

PAGE 100

CHAPTER 8 STUDY DESIGN Introduction The aim of conducting a bioequivalence study is to determine whether the generic formulation could be switched for the reference or branded formulation. This is to ensure that the generic formulation has similar efficacy and safety as that of brand product. If the generic and brand formulation have similar rate and extent of drug availability at th e site of action. We saw that this may not be always possible (feasible ) for example in case of inhaled medications. Alternatively one may argue since they need to be equally effective, one can conduct an efficacy comparison study. While this may be possible with certain class of inhaled medications like bronchodilators, it is not so for inhaled glucocorticoids. Sh allow dose response curve and individual corticosteroid sensitivity are the main r easons that disable use of such efficacy studies. Consequently one needs to look for other tools. We realize that the main questions that need to be answered are as follows How much is deposited in the lung? How much is absorbed orally (only relevant for drugs with significant oral bioavailability)? Where is it deposited (cen tral vs. peripheral lung)? How fast is it absorbed? If two inhalers deliver similar amounts of drug with similar regional distribution as well as similar drug dissolution and absorption profiles, we could say that these in halers are equivalent. Of course one needs to ensure that the inhalers have similar safety profiles as well. We saw in the previous chapters that the pharmacokinetic (PK) tools have the ability to answer these questions relevant to establishi ng BE of inhaled corticosteroids (ICS). The systemic exposure of the drug could be used to estimate systemic sa fety of the inhaled drug. We also saw that in vitro tests such as using cascade impactor data may be needed to answer certain special scenarios (that 100

PAGE 101

may be theoretically possible) or for some dr ugs which are rapidly ab sorbed into systemic circulation (for example budesonide). Bioequivalence study design With all questions being answered we propos e a study design that could be employed to evaluate inhalers for BE. The flow chart is pr esented in figure 8-1 which details the path one could follow to evaluate generic inhalers. To be gin with, one needs to perform a PK study with brand and generic inhalers in he althy volunteers. For dr ugs with significant oral bioavailability (eg. budesonide or beclomethasone), one could em ploy the charcoal block method. The details of using charcoal block method have been discusse d and can also be found in literature [85]. As charcoal adsorbs the drug and prevents its absorp tion from GI tract, one can estimate the oral bioavailability of the drug by administering th e drug with and without charcoal. However most of the newer corticosteroids have negligible oral bioavailab ility (<1%) and hence dont need charcoal block [86-89]. With crossover design, sufficient washout period should be included between various study arms. The AUCs and Cmax should be compared betw een the two inhalers If any of these parameters differ, this means that the inhalers are not equivalent. If they are equivalent then one moves down the decision line to perform further tests. The further tests depend on the type of ICS being compared. Following discussions from previous chapters and the figure 8-1 we demonstrated that if the drug is slowly dissolving such as flu ticasone propionate, similar AUCs would mean similar dose and regional drug depos ition within the lung. We also saw that one may argue that (although pr actically unlikely) theore tically this could also be seen with different regional distribution pattern by adjusting the drug dose from the inhaler. In such specific scenario, one needs to conduct an in vitro study and compare the respirable dose delivered by the two inhalers. If the respirable dos es are not the same, then these i nhalers are not equivalent. With 101

PAGE 102

the respirable does being the same, it can be sa id with sufficient confidence that the dose available and the regional distribution is the same for the two inhalers. This would mean that the two inhalers are bioequivalent. There is another approach th at one could use in this sp ecial scenario. A PK study to compare inhalers in asthmatics can be conducted. We discussed prev iously in chapter 4 that the drug aerosol deposition is more central in asthmatics. The lung congestion causes most of the drug to be deposited in the upper airways. If we had a generic inhaler with more peripheral deposition as compared to brand inhaler then we from run # 42 in table 6-4 we saw that a lower dose (approx 80% of dose compared to the brand inhaler) from generic could give similar AUC as brand inhaler. This scenario had 70% of drug deposited in peripheral areas with generic inhaler. One could only expect su ch high peripheral depos ition with large fraction of particles in 1-3 um range and only in healthy volunteers wi th good airway function. If the same inhalers were then administered to asthmatics we would see a deposition pattern similar to runs # 50-52 (add simulation). With most of the aerosol being deposited in upper airways with both inhalers, a lower dose from generic inhaler would cause most studies to fail when the AUCs are compared. The simulations showed that more than 90% of tria ls fail to show equivalence. Also it is difficult to see such a huge turnaround of regional deposit ion with the same inhaler in healthy and asthmatic subjects. Hence one might compare th e ratio of AUCs in healthy and asthmatics of generic inhaler with that of bra nd inhaler. If they are similar then there are negligible chances that the inhalers differ in the dose delivered and/or regional drug deposition within the lung. If the AUCs and Cmax are same in healthy vo lunteers and the drug is fast dissolving, one can only say that the amount availa ble to the entire lung is the same and the rate at which it is absorbed is same. One still needs to compare the regional drug deposition within the lung. The 102

PAGE 103

respirable dose data from cascade impactor will not be of much help in this case. The drug aerosol particle size determines where exactly wi ll it deposit within th e lung. Smaller particles will tend to have deep (or peripheral) depositi on with larger particles being deposited more centrally. The very large partic les ( >6-10 um) will not enter the lung and be impacted on the throat and swallowed. One needs more detailed an alysis of the cascade imp actor data to have a better comparison of the particle size distribution (PSD) profiles for the two inhalers. If the cascade impactor data suggest differences then, the inhalers are to be cons idered not equivalent. In case of similar PSD profiles, one can conc lude that the inhalers are equivalent. We have presented a detailed study design (f igure 8-1) that could be employed for bioequivalence studies w ith inhaled corticosteroids. This design takes into consideration the lipophilicity of the drug. It also has a ppropriate tests recommended whether in vitro or in vivo in asthmatic population. We see that the PK study needs a single dose and then measuring plasma levels over 24 hours. This is to be followed by a washout period and then administering the other inhaler. In certain cases, a charcoal block arm may be included in the study. The safety profile of these inhalers can be estimated dir ectly from the systemic exposure. This modeling and simulation approach has helped us test the feasibility of using PK tools to evaluate inhaler BE. Such an approach help s understand the trial ou tcomes before actually conducting the trial. Various scenarios can be simulated in silico by using pharmaco-statistical models. These models include all relevant inform ation from pre-clinical and clinical studies. Such models utilize information such as plasma concentration, drug effects and side effects, patient characteristics and other information whic h may lead to better prediction of the study outcome. Better understanding of these outcomes thr ough such models can help improve clinical trial designs. Data from well desi gned trials could in turn help develop models with greater 103

PAGE 104

predictive power. This learn and confirm cycle ha s been described in literature [164]. This kind of model based drug development has been ga ining popularity in the industry [165, 166]. The probability of success of a trial can be estima ted using these pharmacometric simulations and help in making go/no-go decisions in drug develo pment programs. The approach saves time and resources by avoiding unnecessary trials and o ffers better chances of success by better trial design. This approach has also b een supported by the FDA through the Critical Path Initiative [167]. We demonstrated that the PK appro ach in conjunction with appropriate in vitro tests can be effectively used to establish the BE of ICS. Such an approach needs few healthy subjects (30-50) and can be completed over a period of days. It is better suited to de tect differences in the inhalers as compared to the lengthy and inconclusive efficacy trials. The ultimate aim of any bioequivalence test is to ensure that when aswitch is made be tween the generic and the brand product, the patient receives the sim ilar efficacy and safety from use of either product. For this to be possible the test must be able to disti nguish between two products. In case of inhaled corticosteroids, the PK approach possesses the required potential. 104

PAGE 105

Figure 8-1. Study design to conduct bioequivalence trials for inhaled corticosteroids 105

PAGE 106

APPENDIX A NONMEM CODE FOR PK MODEL OF FP AFTER INHALATION $PROBLEM FP PK $INPUT ID ACTI TIME AMT MDV DV EVID CMT $DATA fpdata.csv IGNORE=C $SUBROUTINE ADVAN6 TRANS1 TOL=3 $MODEL COMP=(DEPOT,DEFDOSE);LUNGS COMP=(CENTRAL);PLASMA COMP=(PERIPH);PERIPHERAL $PK TVF1=THETA(1) F1=TVF1*EXP(ETA(1)) TVCL=THETA(2) CL=TVCL*EXP(ETA(2)) TVVC=THETA(3) VC=TVVC*EXP(ETA(3)) K20=TVCL/TVVC K12=THETA(4)*EXP(ETA(4)); K32=THETA(5)*EXP(ETA(5)) TVK23=(K12+THETA(6)); K23=TVK23*EXP(ETA(6)) SC=VC; OUTPUT IN ng/ml $ERROR IPRED=F IRES=DV-IPRED DEL=0 IF (IPRED.EQ.0) DEL=1 IWRE=(1-DEL)*IRES/(IPRED+DEL) Y=F+F*ERR(1); $DES DADT(1)=-K12*A(1) DADT(2)=K12*A(1)-K20*A(2)-K23*A(2)+K32*A(3) DADT(3)=K23*A(2)-K32*A(3) $THETA (0.05,0.3,0.6);Pulmonary deposition (70 FIXED);CL (25 FIXED);VC (0.01,0.1,2);K12 (0.01,0.1,10);K32 (0.01,0.1,10);K23 106

PAGE 107

$OMEGA (0.01);INH (0.01);CL (0.01);VC (0.01);K12 (0.01);K32 (0.01);K23 $SIGMA (0.04); $ESTIMATION METHOD=1 SIGDIGITS=3 INTERACTION MAXEVAL=9999 PRINT=0 POSTHOC $TABLE ID TIME EVID CMT IPRED IWRE IRES CL VC K20 F1 NOPRINT ONEHEADER FILE=sdtabfp1 107

PAGE 108

APPENDIX B R CODE FOR AUC AND CMAX CALCULATION area<-read.table("sdtabpr ob10",header=F,sep=' ') names(area)<-c("ISIM", "ID" ,"TIME", "EVID", "CMT", "OCC", "IPRED", "DV") head(area) area$NEW<-(area$ISIM*1000+area$ID) area$PROD<-(area$DV*area$TIME) area0<-area[area$OCC==0 & area$ CMT==3 &area$EVID==0 ] head(area0) isim0<-area0[!duplicated(area0$NEW), ] tail(isim0) area1<-area[area$OCC==1 & area$ CMT==3 &area$EVID==0 ] head(area1) isim1<-area1[!duplicated(area1$NEW), ] tail(isim1) AUC <-function (data, time = "TIME", id = "ID", dv = "DV") auc0<-AUC(data=area0,id="NEW",time="TIME",dv="DV") auc0$loggen0<-log10(auc0$AUC*10) head(auc0) tail(auc0) auc1<-AUC(data=area1, id="NEW",time="TIME",dv="DV") auc1$loggen1<-log10(auc1$AUC*10) head(auc1) tail(auc1) aucmerge<-merge(auc0,auc1, by="NEW") aucmerge$aucdiff<-(aucmerge$loggen0-aucmerge$loggen1) head(aucmerge) tail(aucmerge) Tmax<-function (data, id = "ID", dv = "DV", time = "TIME") Cmax<-function (data, id = "I D", dv = "DV", time = "TIME", isim="ISIM", new="NEW") tmax0<-Tmax(data=area0, id="NEW",time="TIME",dv="DV") tmax1<-Tmax(data=area1, id="NEW",time="TIME",dv="DV") cmax0<-Cmax(data=area0, id="NEW",t ime="TIME",dv="DV", isim="ISIM") head(cmax0) tail(cmax0) cmax1<-Cmax(data=area1, id="NEW",t ime="TIME",dv="DV", isim="ISIM") 108

PAGE 109

head(cmax1) tail(cmax1) cmerge<-merge(cmax0,cmax1,by="NEW") head(cmerge) tail(cmerge) cmerge$cratio<(log10(cmerge $DV.x)-log10(cmerge$DV.y)) head(cmerge) tail(cmerge) AUMC <-function (data, time = "TIM E", id = "ID", dv = "PROD") aumc0<-AUMC(data=area0,id="NEW",time="TIME",dv="PROD") aumc1<-AUMC(data=area1, id="NEW",time="TIME",dv="PROD") aumcmerge<-merge(aumc0,aumc1,by="NEW") bigdata<-merge(aumcmerge,aucmerge, by="NEW") set<-merge(bigdata,cmerge, by="NEW") set$mrt0<-(set$AUMC.x/set$AUC.x) set$mrt1<-(set$AUMC.y/set$AUC.y) Myfunc3
PAGE 110

LIST OF REFERENCES 1. Weibel ER (1963) Morphometry of the Human Lung. Springer Academic Press, New York 2. Hickey AJ and Thompson DC (2004) Physio logy of the Airways. Marcel Dekker, New York 3. Godfrey RW (1997) Human airway epithelia l tight junctions. Microscopy Research and Technique 38: 488-499 4. Stuart BO (1976) Deposition and clearance of inhaled particles. Environ Health Perspect 16: 41-53 5. Schanker LS, Mitchell EW and Brown RA Jr (1986) Species comparison of drug absorption from the lung after aerosol inha lation or intratrach eal injection. Drug Metabolism and Disposition 14: 79-88 6. Brown RA Jr and Schanker LS (1983) Abso rption of aerosolized drugs from the rat lung. Drug Metabolism and Disposition 11: 355-60 7. Edsbcker S, Wollmer P, Selroos O, Borg strm L, Olsson B and Ingelf J (2008) Do airway clearance mechanisms influence the local and systemic effects of inhaled corticosteroids? Pulmonary Pharm acology and Therapeutics 21: 247-258 8. Stuart B (1984) Deposition and Clearance of Inhaled Particles. Environment Health Perspectives 55: 369-390 9. Currie DC, Pavia D, Agnew JE, Lopez-Vidrie ro MT, Diamond PD, Cole PJ and Clarke SW (1987) Impaired tracheobronchial cleara nce in bronchiectasis. Thorax 42: 126-130 10. Byron PR (1986) Prediction of drug residence times in regions of the human respiratory tract following aerosol inhalation. Journa l of Pharmaceutical Sciences 75: 433-438 11. Derom E (2003) Pulmonary Deposition and Effects of Aerosolized Drugs in Pulmonary Patients. Kluwer Academic Publishers, Dordecht, The Netherlands 110

PAGE 111

12. Bateman J R, Pavia D, Sheahan N F, Agnew J E and Clarke S W (1983) Impaired tracheobronchial cleara nce in patients with mild st able asthma. Thorax 38: 463-467 13. Lippmann M, Leikauf G, Spektor D, Schlesinger RB and RE A. (1981) The effects of irritant aerosols on mucus clearance from large and small conductive airways. Chest 80: 873-7 14. Weiss T, Dorow P and Felix R (1981) Effects of a beta adrenergic drug and a secretolytic agent on regional mucociliary clearance in patients with COLD. Chest 80: 881-5 15. Hasani A, Agnew JE, Pavia D, Vora H and Clarke SW (1993) Effect of oral bronchodilators on lung mucociliary clearan ce during sleep in patients with asthma. Thorax 48: 287-9 16. Sutton PP, Pavia D, Bateman JR and Clarke SW (1981) The effect of oral aminophylline on lung mucociliary clearan ce in man. Chest 80: 889-92 17. Global Initiative for Asthma (1995) Globa l Strategy for asthma management and prevention, NHLBI/WHO workshop, National Heart, Lung and Blood Institute. 1-176 18. Pauwels R, Newman S and Borgstrom L (1997) Airway deposition and airway effects of antiasthma drugs delivered from metered-dose inhalers. European Re spiratory Journal 10: 2127-2138 19. National Heart Lung and Blood Institute, Di sease and Conditions Index, [Updated Feb 1 2009, cited Feb 15], Available from: http://www.nhlbi.nih.gov/health/dci/D iseases/Asthma/Asthma_WhatIs.html 20. Barnes PJ and Godfrey S (1998) Asthma Therapy. Martin Dunitz, Malden 21. Solway J and Fredberg JJ (1997) Perhaps airway smooth muscle dysfunction contributes to asthmatic bronchial hyperresponsiveness af ter all. American Journal of Respiratory Cell amd Molecular Biology 17: 144-146 22. Devalia JL and Davies RJ (1993) Airway epithelial cells and mediators of inflammation. Respiratory Medicine 87: 405-408 111

PAGE 112

23. Barnes PJ (1998) Pathophysiology of Asthma. Academic Press, San Diego 24. Azzawi M, Bradley B, Jeffery PK, Frew AJ Wardlaw AJ, Knowles G, Assoufi B, Collins JV, Durham S and Kay AB (1990) Identific ation of activated T lymphocytes and eosinophils in bronchial biopsies in stab le atopic asthma. Am erican Review of Respiratory Disease 142: 1407-1413 25. Dunnill MS (1960) The pathology of asthma, with special reference to changes in the bronchial mucosa. Journal of Clinical Pathology 13: 27-33 26. Dunnill MS, Massarella GR and Anderson JA (1969) A comparison of the quantitative anatomy of the bronchi in normal subjects, in status asthmaticus, in chronic bronchitis and in emphysema. Thorax 24: 176-179 27. Hamid Q, Song Y, Kotsimbos TC, Minshall E, Bai TR, Hegele RG and Hogg JC (1997) Inflammation of small airways in asthma. Journal of Allergy a nd Clinical Immunology 100: 44-51 28. Vignola AM, Chanez P, Campbell AM, Souques F, Lebel B, Enander I and Bousquet J (1998) Airway Inflammation in Mild Intermittent and in Persistent Asthma American Journal of Respiratory and Critical Care Medicine 157: 403-409 29. Barnes PJ and Adcock IM (2003) How do corticosteroids work in asthma? Annals of Internal Medicine 139: 359-370 30. Barnes PJ (2005) How corticosteroids cont rol inflammation: Quintiles Prize Lecture 2005. British Journal of C lin Pharmacology 148: 245-254 31. Barnes PJ and Karin M (1997) Nuclear facto r-kappaB: a pivotal tran scription factor in chronic inflammatory diseases. The New E ngland Journal of Me dicine 336: 1066-1071 32. Barnes PJ and Adcock IM (1998) Transcription factors and asthma. European Respiratory Journal 12: 221-234 33. Barnes PJ, Pedersen S and Busse WW (1998) Efficacy and safety of inhaled corticosteroids. American Journal of Respir atory and Critical Care Medicine 157: S1-S53 112

PAGE 113

34. Wu K, Goyal N, Stark JG and Hochhaus G (2008) Evaluation of the administration time effect on the cumulative cortisol suppressi on and cumulative lymphocytes suppression for once-daily inhaled corticosteroids: a population modeling/simulation approach. Journal of Clinical Pathology 48: 1069-1080 35. Thorsson L E. S., Conradson TB (1994) L ung deposition of budesonide from Turbuhaler is twice that from a pressurized metered-dos e inhaler P-MDI. The European Respiratory Journal 7: 1839-1844 36. Ryrfeldt A A. P., Edsbcker S, Tnnesson M, Davies D, Pauwels R (1982) Pharmacokinetics and metabolism of budesoni de, a selective gluc ocorticoid. European Journal of Respiratory Diseases 122: 86-95 37. Hochhaus G (2004) New Developments in Corticosteroids. The Proceedings of the American Thoracic Society 1: 269-274 38. Crompton GK (1982) Problems patients have using pressurized aerosol inhalers. European Journal of Respiratory Diseases 63: 101-104 39. Goodman DE, Israel E, Rosenberg M, J ohnston R, Weiss ST and Drazen JM (1994) The influence of age, diagnosis, and gender on proper use of metered-dose inhalers. American Journal of Critical Care in Medicine 150: 1256-1261 40. Thorsson L and Edsbcker S (1998) Lung depos ition of budesonide from a pressurized metered-dose inhaler attached to a spacer. European Respiratory Journal 12: 1340-1345 41. Agertoft L and Pedersen S (1993) Importa nce of inhalation devi ce on the effect of budesonide. Archives of Disease in Childhood 69: 130-133 42. Langdon CG and Thompson J (1994) A multicentre study to compare the efficacy and safety of inhaled fluticasone propionate a nd budesonide via metered-dose inhalers in adults with mild-to-moderate asthma. Britis h journal of Clinical Research 5: 73-84 43. Engel T, Heinig JH, Malling HJ, Scharling B, Nikander K and Madsen F (1989) Clinical comparison of inhaled budesonide delivered either via pressurized metered dose inhaler or Turbuhaler. Allergy 44: 220-225 113

PAGE 114

44. Borgstrm L, Bondesson E, Morn F, Trof ast E and Newman SP (1994) Lung deposition of budesonide inhaled via Turbuhaler: a compar ison with terbutaline sulphate in normal subjects. European Respir atory Journal 7: 69-73 45. Nave R, Zech K and Bethke T (2005) Lo wer oropharyngeal de position of inhaled ciclesonide via hydrofluoroalk ane metered-dose inhaler comp ared with budesonide via chlorofluorocarbon metered-dose inhaler in healthy subjects. European Journal of Clinical Pharmacology 61: 203-208 46. Chrystyn H (1994) Standards for Bioequi valence of Inhaled Products. Clinical Pharmacokinetics 26: 1-6 47. Polosa R and Holgate ST (1997) Adenosine bronchoprovocation: a promising marker of allergic inflammation in asthma? Thorax 52: 919-923 48. Sterk PJ, Fabbri LM, Quanjer PH, Cockcroft DW, O'Byrne PM, Anderson SD, Juniper EF and Malo JL (1993) Airway responsivene ss. Standardized challenge testing with pharmacological, physical and sensitizing s timuli in adults. Report Working Party Standardization of Lung Function Tests, European Community for Steel and Coal. Official Statement of the European Respirat ory Society. European Respiratory Journal 6: 53-83 49. Ahrens RC, Hendeles L, Clarke WR, Dockhorn RJ, Hill MR, Vaughan LM, Lux C and Han SH (1999) Therapeutic equivalence of Spiros dry powder i nhaler and Ventolin metered dose inhaler. A bioassay using methach oline. American Journal of Respiratory and Critical Care Medicine 160: 1238-1243 50. Juniper EF, Frith PA, Dunnett C, Cockcroft DW and Hargreave FE (1978) Reproducibility and comparison of responses to inhaled histamine and methacholine. Thorax 33: 705-710 51. Currie GP, Bates CE, Lee DK, Jackson CM and Lipworth BJ (2003) Effects of fluticasone plus salmeterol versus twice th e dose of fluticasone in asthmatic patients. European Journal of Clinical Pharmacology 59: 11-15 52. Sont JK, Willems LN, Bel EH, van Krieken JH, Vandenbroucke JP and Sterk PJ (1999) Clinical control and histopathologic out come of asthma when using airway 114

PAGE 115

hyperresponsiveness as an additional guide to long-term treatment. The AMPUL Study Group. American Journal of Respiratory and Critical Care Medicine 159: 1043-1051 53. Van Den Berge M, Meijer RJ, Kerstjens HA, de Reus DM, Koter GH, Kauffman HF and Postma DS (2001) PC(20) adenosine 5' -monophosphate is more closely associated with airway inflammation in asthma than PC(20) methacholine. American Journal of Respiratory and Critical Care Medicine 163: 1546-1550 54. O'Connor BJ, Ridge SM, Barnes PJ and Fu ller RW (1992) Greater effect of inhaled budesonide on adenosine 5'-monophosphate-ind uced than on sodium-metabisulfiteinduced bronchoconstriction in asthma. Ameri can Review of Respiratory Disease 146: 560-564 55. Taylor DA, Jensen MW, Kanabar V, Engels ttter R, Steinijans VW, Barnes PJ and O'Connor BJ (1999) A Dose-dependent Effect of the Novel Inhaled Corticosteroid Ciclesonide on Airway Responsiveness to Adenosine-5'-Monophosphate in Asthmatic Patients American Journal of Respirator y and Critical Care Medicine 160: 237-243 56. Luijk B, Kempsford RD, Wright AM, Za nen P and Lammers JW (2004) Duration of effect of single-dose inhaled flut icasone propionate on AMP-induced bronchoconstriction. European Re spiratory Journal 23: 559-564 57. Kharitonov SA, Yates DH, Chung KF and Barnes PJ (1996) Changes in the dose of inhaled steroid affect exhale d nitric oxide levels in asthmatic patients. European Respiratory Journal 9: 196-201 58. Kharitonov SA and Barnes PJ (2000) Clinical Aspects of Exhaled Nitric Oxide. European Respiratory Journal 16: 781-792 59. Kharitonov SA, Yates D, Robbins RA, Logan-Si nclair R, Shinebourne EA and Barnes PJ (1994) Increased nitric oxide in exhaled ai r of asthmatic patients. Lancet 343: 133-135 60. Smith AD, Cowan JO, Filsell S, McLachlan C, Monti-Sheehan G, Jackson P and Taylor DR (2004) Diagnosing asthma: comparisons be tween exhaled nitric oxide measurements and conventional tests. American Journal of Respiratory and Criti cal Care Medicine 169 61. Szefler SJ, Mitchell H, Sorkness CA, Gergen PJ, O'Connor GT, Morgan WJ, Kattan M, Pongracic JA, Teach SJ, Bloomberg GR, Eggleston PA, Gruchalla RS, Kercsmar CM, 115

PAGE 116

Liu AH, Wildfire JJ, Curry MD and Busse WW (2008) Management of asthma based on exhaled nitric oxide in additi on to guideline-base d treatment for inner-city adolescents and young adults: a randomised cont rolled trial. Lancet 372: 1065-1072 62. Shaw DE, Berry MA, Thomas M, Green RH Brightling CE, Wardla w AJ and Pavord ID (2007) The use of exhaled nitric oxide to guide asthma management: a randomized controlled trial. American Journal of Resp iratory and Critical Care Medicine 176: 231237 63. Newman SP, Wilding IR and Hirst PH ( 2000) Human lung deposition data: the bridge between in vitro and clinical evaluations for inhaled drug products? Internationa l Journal of Pharmaceutics 208: 49-60 64. Zanen P and Lammers JWJ (1995) Sample sizes for comparative inhaled corticosteroid trials with emphasis on showing therapeutic eq uivalence. European Journal of Clinical Pharmacology 48: 179-184 65. Edsbacker S and Szefler S (1997) Glucocorti coid Pharmacokinetics. Marcel Dekker Inc, New York 66. Pedersen S and Hansen OR (1995) Budesonide treatment of moderate and severe asthma in children: a dose-response study. Journal of Allergy and Clinical Immunology 95: 2933 67. Pearlman D, Noonan M, Tashkin D, Gold stein M, Hamedani A, Kellerman D and Schaberg A (1997) Comparative efficacy and safety of twice daily FP powder Vs Placebo in treatment of moderate asthma. Annals of Allergy, Asthma and Immunology 78: 356362 68. Singh D, Tutuncu A, Lohr I, Carlho lm M and Polanowski T (2007) Budesonide administered using chlorofluorocarbon and hydrofluoroalkane pressurized metered-dose inhalers: pharmacokinetics, pharmacodynamics and clinical equivale nce. International Journal of Clinical Pharmacology and Therapeutics 45: 485-495 69. Peden D, Berger W, Noonan M, Thomas M, Hendricks V, Hamedani A, Mahajan P and House K (1998) Inhaled fluti casone propionate delivered by means of two different multidose powder inhalers is effective and safe in a large pediatric population with persistent asthma. Journal of Alle rgy and Clinical Immunolology 102: 32-38 116

PAGE 117

70. Stewart BA, Ahrens RC, Carrier S, Frosolono M, Lux C, Han SH and Milavetz G (2000) Demonstration of in vivo bioequivalence of a generic albuterol mete red-dose inhaler to Ventolin. Chest 117: 714-721 71. Ahrens RC, Teresi ME, Han SH, Donnell D, Vanden Burgt JA and Lux CR (2001) Asthma Stability after Oral Prednisone. Amer ican Journal of Respiratory Critical Care Medicine 164: 1138-1145 72. Ahrens R. (1991) On comparing -adrenerg ic agonists. Annals of Allergy 67: 296-298 73. Ahrens RC H. J., Milavetz G, Annis L, Ri es R (1987) Use of bronchial provocation with histamine to compare the pharmacodynamics of inhaled albuterol and metaproterenol in patients with asthma. Journal of Alle rgy and Clinical Immunology 79: 876-882 74. Ahrens R., Nelson, JS, Lux, C, et al. (1992) A method for determining bioequivalence of -agonist metered-dose inhalers. Journal of Allergy and Clinical Immunology 89 (Suppl): 340 75. Harris JB A. R., Milavetz G, et al. (1986) Relative potencies and rates of decline in effect of inhaled albuterol (A) and terbutaline (T). Journal of Allergy and Clinical Immunology 77 (Suppl): 147 76. Finney D. (1978) Statistical methods in bi ological assay. Charles Griffin & Co, London, UK 77. Stewart BA A. R., Carrier S, Frosol ono M, Lux C, Han SH, Milavetz G (2000) Demonstration of in vivo bioequivalence of a generic albuterol mete red-dose inhaler to Ventolin. Chest 117: 714-721 78. Ahrens RC T. M., Han SH, Donnell D, Vanden Burgt JA, Lux CR (2001) Asthma Stability after Oral Prednisone. American Jour nal of Respiratory Crit ical Care Medicine 164: 1138-1145 79. Barnes PJ P. S., Busse WW (19998) Efficacy and safety of inhaled corticosteroids: New developments. American Journal of Respirat ory and Critical Care Medicine 157: S1-53 117

PAGE 118

80. Pedersen S O. B. P. (1997) A comparis on of the efficacy and safety of inhaled corticosteroids in asthma. Allergy 52: 1-34 81. Krishnaswami S, Mllmann H, Derendorf H and Hochhaus G (2000) A sensitive LCMS/MS method for the quantification of fluti casone propionate in human plasma. journal of Pharmaceutical and Biomedical Analysis 22: 123-129 82. Dimova H, Wang Y, Pommery S, Moellm ann H and Hochhaus G (2003) SPE/RIA vs LC/MS for measurement of low levels of budesonide in plasma. Biomedical Chromatography 17: 14-20 83. Singh SD, Whale C, Houghton N, Daley-Yate s P, Kirby SM and Woodcock AA (2003) Pharmacokinetics and systemic effects of inhaled fluticasone propionate in chronic obstructive pulmonary disease. British Journal of Clin ical Pharmacology 55: 375-381 84. Daley-Yates PT, Price AC, Sisson JR, Pereira A and Dallow N (2001) Beclomethasone dipropionate: absolute bi oavailability, pharmacokinetics and metabolism following intravenous, oral, intranasal and inhaled administration in man. British Journal of Clinical Pharmacology 51: 400-409 85. Thorsson L, Edsbacker S and Conradson TB (1994) Lung deposition of budesonide from Turbuhaler is twice that from a pressurized metered-dose inhaler P-MDI. The European Respiratory Journal 7: 1839-1844 86. Mobley C and Hochhaus G (2001) Pharmac okinetic considerations in the design of pulmonary drug delivery systems for gluc ocorticoids. Marcel Decker, New York 87. Hochhaus G, Sahasranaman S, Derendorf H and Moellmann H (2002) Intranasal glucocorticoid delivery: competition between local and systemic effects. STP Pharma Sciences, 12: 23-31 88. Hochhaus G, Derendorf H, Talton J and M llmann H (2002) Factors involved in the pulmonary targeting of inhaled glucoc orticoids. Marcel Dekker, New York 89. Kelly HW (2005) Mometasone Furoate Inhalation Powder. Pedriatics Asthma, Allergy and Immunology 18: 167-169 118

PAGE 119

90. Lhelm S, Kirjavainen M, Kela M, Herttu ainen J, Vahteristo M, Silvasti M and RankiPesonen M (2005) Equivalent l ung deposition of budesonide in vivo: a comparison of dry powder inhalers using a pharmacokinetic method. British Journal of Clinical Pharmacology 59: 167-173 91. Daley-Yates P, Parkins D, Thomas M, Gillett B, House K and Ortega H (2009) Pharmacokinetic, Pharmacodynamic, Efficacy, and Safety Data From Two Randomized, Double-Blind Studies in Patients With Asth ma and an In Vitro Study Comparing Two Dry-Powder Inhalers Delivering a Combina tion of Salmeterol 50 g and Fluticasone Propionate 250 g: Implications for Establ ishing Bioequivalence of Inhaled Products. Clinical Therapeutics 31: 370-385 92. Harrison TW and Tattersfield AE (2003) Plasma concentrations of fluticasone propionate and budesonide following inhalation from dr y powder inhalers by healthy and asthmatic subjects. Thorax 58: 258-260 93. Harrison TW, Wisniewski A, Honour J and Tattersfield AE (2001) Comparison of the systemic effects of fluticasone propionate a nd budesonide given by dry powder inhaler in healthy and asthmatic subj ects. Thorax 56: 186-191 94. Brutsche MH, Brutsche IC, Munawar M, Langley SJ, Masterson CM, Daley-Yates PT, Brown R, Custovic A and Woodcock A (2000) Comparison of pharmacokinetics and systemic effects of inhaled fluticasone propi onate in patients with asthma and healthy volunteers: a randomised cross over study. Lancet 356: 556-561 95. Derom E and Pauwels R (1995) Bioequivalence of inhaled drugs. European Respiratory Journal 8: 1634-1636 96. Borgstrm L and Nilsson M (1990) A met hod for determination of the absolute pulmonary bioavailability of inhaled drugs: terbutaline. Pharmaceutical Research 7: 1068-1070 97. Borgstrm L, Newman S, Weisz A and Mo rn F (1992) Pulmonary deposition of inhaled terbutaline: comparison of scanning gamma camera and urinary excretion methods. Journal of Pharmaceutical Sciences 81: 753-755 98. Howarth PH (1997) What is the nature of as thma and where are the therapeutic targets? Respiratory Medicine 91: 2-8 119

PAGE 120

99. Newman SP (1993) Scintigraphic assessment of therapeutic aerosols. Critical Review of Therapeutic Drug Carrier Systems 10: 65-109 100. Chrystyn H (2000) Methods to determine l ung distribution of inhaled drugs could gamma scintigraphy be the gold standard? British Journal of Clinical Pharmacology 49: 525-528 101. Newman SP (2000) Can lung de position data act as a surrogate for the clinical response to inhaled asthma drugs? British Journal of Clinical Pharmacology 49: 529-537 102. Fleming JS, Nassim M, Hashish AH, Bailey AG, Conway J, Holgate S, Halson P, Moore E and Martonen TB (1995) Description of Pulmonary Deposition of Radiolabeled Aerosol by Airway Generation Using a C onceptual 3-Dimensional Model of Lung Morphology. Journal Of Aerosol Medicine 8: 341-356 103. Heald DL, Berridge MS, Lee Z, Leisure GP, Dalby RN, Byron PR and Farr SJ (1998) In vivo deposition of triamcinolone acetonide (A zmacort) with and without the integrated spacer. Respiratory Drug Delivery VI Buffalo Grove: Interpharm Press 345-347 104. Selroos O, Pietinalho A and Riska H (1996) Delivery devices for inhaled asthma medication. Clinical Immunotherapy 6: 273-299 105. Newman SP, Millar AB, Lennard-Jones TR, Morn F and Clarke SW (1984) Improvement of pressurised aerosol depositi on with Nebuhaler spacer device. Thorax 39: 935-941 106. Thorsson L, Kenyon CJ, Newman SP a nd Borgstrm L (1998) Lung deposition of budesonide in asthmatics: a comparison of diffe rent formulations. International Journal of Pharmaceutics 168: 119-127 107. Toogood JH, Baskerville J, Jennings B, Lefc oe NM and Johansson S-A (1984) Use of spacers to facilitate inhaled corticosteroid treatment in asthma. American Review of Respiratory Diseases 129: 723-729 108. Newman SP and Newhouse MT (1996) Effect of add-on devices for aerosol drug delivery: deposition studies and clinical aspects. Journal Of Aerosol Medicine 9: 55-70 120

PAGE 121

109. Lfdahl CG, Andersson L, Bondesson E, Carl sson LG, Friberg K, Hedner J, Hrnblad Y, Jemsby P, Klln A, Ullman A, Werner S and Svedmyr N (1997) Differences in bronchodilating potency of salbutamol in Tur buhaler as compared with a pressurized metered-dose inhaler formulation in patie nts with reversible airway obstruction. European Respiratory Journal 10: 2474-2478 110. Borgstrm L, Derom E, Sthl E, Whlin -Boll E and Pauwels R (1996) The inhalation device influences lung deposition and bronchod ilating effect of ter butaline. American Journal of Respiratory and Criti cal Care Medicine 153: 1636-1640 111. Borgstrm L, Olsson B and Thorsson L (2006) Degree of throat deposition can explain the variability in lung deposition of inhaled drugs. Journal Of Aerosol Medicine 19: 473483 112. Borgstrm L, Bengtsson T, Derom E and Pa uwels R (2000) Variability in lung deposition of inhaled drug, within and be tween asthmatic patients, w ith a pMDI and a dry powder inhaler, Turbuhaler. International Journal of Pharmaceutics 193: 227-230 113. Minto C, Li B, Tattam B, Brown K, Seale JP and Donnelly R (2000) Pharmacokinetics of epimeric budesonide and fluticasone propionate after repeat dose inha lation intersubject variability in systemic absorption from the l ung. British Journal of Clinical Pharmacology 50: 116-124 114. Snell N (1998) Equivalence testingth e special case of inhaled medications. International Journal of Pharmaceutical Medicine 12: 245-246 115. Webb J (1998) Avoidance of regulatory delays for inhalation products: alternative approaches to product development and regulat ory approval. Interpharm Press, Buffalo Grove 116. Clark AR, Gonda I and Newhouse MT (1998) Towards meaningful laboratory tests for evaluation of pharmaceutical aerosols. Journal Of Aerosol Medicine 11: S1-S7 117. Hickey AJ (1992) Methods for aerosol par ticle size characterisation. Marcel Dekker, New York 121

PAGE 122

118. Martonen TB, Katz I, Fults K and Hickey AJ (1992) Use of analytically defined estimates of aerosol respirable fraction to predic t lung deposition patterns. Pharmaceutical Research 9: 1634-1639 119. Martonen TB and Katz I (1993) Deposition patt ers of polydisperse aerosols within human lungs. Journal Of Aerosol Medicine 6: 251-274 120. Rogers DF and Ganderton D (1995) Determ ining equivalence of inhaled medications. Respiratory Medicine 89: 253-261 121. Adams WP (1999) Bioavailability and Bioe quivalence Studies for Nasal Aerosols and Nasal Sprays for Local Action Draft Guidan ce. FDA, Center for Drug evaluation and Research 122. Stein SW (2005) 'Squeezing Blood Out of a Turnip'Why Pharmaceutical Industry Struggles So Much with Cascade Impactor Data. 123. Naini V, Chaudhry S, Berry J, Sharpe S, Hart J and Sequeira J (2004) Entry port selection for detecting particle size differences in me tered dose inhaler formulations using cascade impaction. Drug Delivery and Industrial Pharmacy 30: 75-82 124. Stein SW and Myrdal PB (2004) A theoretical and experimental anal ysis of formulation and device parameters affecting soluti on MDI size distributions. Journal of Pharmaceutical Sciences 93: 2158-2175 125. Health Canada (2007-08-01) Submission Re quirements for Subsequent Market Entry Inhaled Corticosteroid Products for Use in the Treatment of Asthma. 126. Boulet LP, Cockcroft DW, Toogood J, Laca sse Y, Baskerville J and Hargreave FE (1998) Comparative assessment of safety and e fficacy of inhaled corticosteroids: report of a committee of the Canadian Thoracic So ciety. European Respiratory Journal 11: 1194-1210 127. Parmeswaran K, Leigh R, O'Byrne PM, Kelly MM, Goldsmith CH, Hargreave FE and M. D. (2003) Clinical models to compare the safety and efficacy of inhaled corticosteroids in patients with asthma. Canadian Respiratory Journal 10: 27-34 122

PAGE 123

128. Committee for Medicinal Products for Hu man Use (CHMP) (2009) Guideline on the requirements for clinical documentation for or ally inhaled products (oip) including the requirements for demonstration of therapeutic equivalence between two inhaled products for use in the treatment of asthma and chronic obstructive pulmona ry disease (cop) in adults and for use in the treatment of asthma in children and adolescents. 129. EMEA (2007) Guideline on the requirements for clinical documentation for Orally inhaled products (OIP) incl uding the requirements for Demo nstration of therapeutic equivalence between two inhaled Products for use in the treatment of asthma and chronic Obstructive pulmonary disease (C OPD). CPMP/EWP/4151/00 Rev 1: 130. Food and Drug Administration (2003) Draft gui dance for industry bioavailability and bioequivalence studies for nasal aerosols and nasal sprays for local action. 131. FDA Center for Drug Evaluation and Resear ch (2000) Summary minutes of the Orally Inhaled and Nasal Drug Products (OINDP) subcommittee of Advisory Committee for Pharmaceutical Sciences. 132. Adams WP, Christopher D, Lee DS, Mo rgan B, Pan Z, Singh GJ, Tsong Y and Lyapustina S (2007) Product Quality Research Institute evaluation of cascade impactor profiles of pharmaceutical aero sols, part 1: background for a statistical method. AAPS PharmSciTech 8: E1-E6 133. Christopher D, Adams WP, Lee DS, Morgan B, Pan Z, Singh GJ, Tsong Y and Lyapustina S (2007) Product Quality Research Institute evaluation of cascade impactor profiles of pharmaceutical aero sols: part 2--evaluation of a method for determining equivalence. AAPS PharmSciTech 8: 134. Christopher D, Adams W, Amann A, Bertha C, Byron PR, Doub W, Dunbar C, Hauck W, Lyapustina S, Mitchell J, Morgan B, Ni chols S, Pan Z, Singh GJ, Tougas T, Tsong Y, Wolff R and Wyka B (2007) Product Quality Research Institute evaluation of cascade impactor profiles of pharmaceutical aerosols. Part 3. Final report on a statistical procedure for determining equiva lence. AAPS PharmSciTech 8: 135. Committee for Proprietary Medicina l Products (1993) Replacement of Chlorofluorocarbons (CFC) in Mete red Dose Inhalation Products. 123

PAGE 124

136. Mllmann H, Wagner M, Krishnaswami S, Dimova H, Tang Y, Falcoz C, Daley-Yates PT, Krieg M, Stckmann R, Barth J, Lawl or C, Mllmann AC, Derendorf H and G H. (2001) Single-dose and steady-state pharm acokinetic and pharmacodynamic evaluation of therapeutically clinically equivalent doses of inhaled fluticasone propionate and budesonide, given as Diskus or Turbohaler dr y-powder inhalers to healthy subjects. Journal of Clinical Pharmacology 41: 1329-1338 137. Mollmann H, Wagner M, Krishnaswami S, Dimova H, Tang Y, Falcoz C, Daley-Yates P, Krieg M, Stockmann R, Barth J, Lawlor C, Mollmann AC, Derendorf H and Hochhaus G (2001) Single-dose and steady-state pharm acokinetic and pharmacodynamic evaluation of therapeutically clinically equivalent doses of inhaled fluticasone propionate and budesonide, given as Diskus or Turbohaler dr y-powder inhalers to healthy subjects. J Clin Pharmacol 41: 1329-1338 138. FDA (2003) Bioavailability and Bioequivalen ce Studies for Orally Administered Drug Products General Considerations. 139. Balthasar J (1999) Bioequivalence and Bioe quivalency Testing. Am erican Journal of Pharmaceutical Education 63: 194-198 140. Krishnaswami S, Hochhaus G, Mllmann H, Barth J and H D. (2005) Interpretation of absorption rate data for inhaled fluticas one propionate obtained in compartmental pharmacokinetic modeling. International J ournal of Clinical Pharmacology and Therapeutics 43: 117-122 141. Brindley C, Falcoz C, Mackie AE and By e A (2000) Absorption kinetics after inhalation of fluticasone propionate via the Diskhaler, Diskus and metered-dose inhaler in healthy volunteers. Clinical Pharmacokinetics 39: 1-8 142. Falcoz C, Oliver R, McDowall JE, Ventresca P, Bye A and Daley-Yates PT (2000) Bioavailability of orally administered micronised fluticasone propionate. Clinical Pharmacokinetics 39: 9-15 143. Harding SM (1990) The human pharmacology of fluticasone propionate. Respiratory Medicine 84: 25-29 124

PAGE 125

144. Nave R, Bethke TD, van Marle SP a nd Zech K (2004) Pharmacokinetics of [14C]ciclesonide after oral and intravenous administration to healthy subjects. Clin Pharmacokinet 43: 479-486 145. Daley-Yates PT, Kunka RL, Yin Y, A ndrews SM, Callejas S and Ng C (2004) Bioavailability of fluticasone propionate a nd mometasone furoate aqueous nasal sprays. Eur J Clin Pharmacol 60: 265-268 146. Ryrfeldt A, Andersson P, Edsbcker S, Tnnesson M, Davies D and Pauwels R (1982) Pharmacokinetics and metabolism of budesoni de, a selective gluc ocorticoid. European Journal of Respiratory Diseases 122: 86-95 147. Hochhaus G, Mollmann H, Derendor f H and Gonzalez-Rothi RJ (1997) Pharmacokinetic/pharmacodynamic aspects of ae rosol therapy using glucocorticoids as a model Journal of Clinical Pharmacology 37: 881-892 148. Camner P, Mossberg B and Philipson K (1973) Tracheobronchial clearance and chronic obstructive lung disease. Scandivian Jour nal of Respiratory Diseases 54: 272-281 149. Lourenco RV, Klimek MF and Borowski CJ (1971) Deposition and clearance of 2 micron particles in the tracheobronchi al tree of normal subjects --smokers and nonsmokers. The Journal of Clinical Investigation 50: 1411-1420 150. Gonda I (1988) Drugs Administered Dirctly into the Respir atory Tract : Modeling of the Duration of Effective Drug Levels. Journa l of Pharmaceutical Sciences 77: 340-346 151. Clark AR and Byron PR (1985) Drug absorp tion from inhalation aerosols administered by positive-pressure ventilation. II : Effect of disodium fluoresce in aerosol particle size on fluorescein absorption kineti cs in the beagle dog resp iratory tract. Journal of Pharmaceutical Sciences 74: 939-942 152. Brown RA Jr and Schanker LS (1983) Abso rption of aerosolized drugs from the rat lung. Drug Metabolism and Disposition 11: 355-360 153. Ryrfeldt A and Bodin NO (1975) The physiologi cal disposition of i buterol, terbutaline and isoproterenol after endot racheal instillation to ra ts. Xenobiotica 5: 521-529 125

PAGE 126

154. Finkbeiner WE (1999) Physiol ogy and pathology of tracheobr onchial glands. Respiratory Physiology 118: 77-83 155. Phipps R J (1981) The airway mucociliary system. International Review of Physiology 23: 213-60 156. Knowles MR and Boucher RC (2002) Mucus clearance as a primary innate defense mechanism for mammalian airway s. J Clin Invest 109: 571-7 157. Harrison TW and Tattersfield AE (2003) Plasma concentrations of fluticasone propionate and budesonide following inhalation from dr y powder inhalers by healthy and asthmatic subjects. Thorax 58: 258-60 158. Mortimer K. J., Harrison T. W., Tang Y., Wu K., Lewis S., Sahasranaman S., Hochhaus G. and Tattersfield A. E. (2006) Plasma concentrations of inhaled corticosteroids in relation to airflow obstruction in as thma. Br J Clin Pharmacol 62: 412-9 159. Mortimer K. J., Tattersfield A. E., Tang Y ., Wu K., Lewis S., Hochhaus G. and Harrison T. W. (2007) Plasma concentrations of fl uticasone propionate and budesonide following inhalation: effect of induced bronchoconstr iction. British Journal of Clin Pharmacology 64: 439-44 160. Pickering H, Pitcairn GR, Hirst PH, Bac on PR, Newman SP, Affrime MB and Marino M (2000) Regional lung deposition of a tec hnetium 99m-labeled formulation of mometasone furoate administered by hydr ofluoroalkane 227 metered-dose inhaler. Clinical Therapeutics 22: 1483-93 161. Saari SM, Vidgren MT, Koskinen MO, Turjanmaa VM, Waldrep JC and Nieminen MM (1998) Regional Lung Deposition and Clearan ce of 99mTc-Labeled BeclomethasoneDLPC Liposomes in Mild and Severe Asthma Chest 113: 1573-9 162. Usmani OS, Biddiscombe MF and Barnes PJ (2005) Regional Lung Deposition and Bronchodilator Response as a Function of {be ta}2-Agonist Particle Size. American Journal of Respiratory and Criti cal Care Medicine 172: 1497-502 126

PAGE 127

163. Mortimer KJ, Harrison TW, Tang Y, Wu K, Lewis S, Sahasranaman S, Hochhaus G and Tattersfield AE (2006) Plasma concentrations of inhaled corticoste roids in relation to airflow obstruction in asthma. Br J Clin Pharmacol 62: 412-419 164. Sheiner LB (1997) Learning ve rsus confirming in clinical drug development. Clinical Pharmacology and Therapeutics 61: 275-291 165. Lalonde RL, Kowalski KG, Hutmacher MM, Ewy W, Nichols DJ, Milligan PA, Corrigan BW, Lockwood PA, Marshall SA, Benincosa LJ, Tensfeldt TG, Parivar K, Amantea M, Glue P, Koide H and Miller R (2007) M odel-based Drug Development. Clinical Pharmacology and Therapeutics 82: 21-32 166. Miller R, Ewy W, Corrigan BW, Ouellet D, Hermann D, Kowalski KG, Lockwood P, Koup JR, Donevan S, El-Kattan A, Li CS, We rth JL, Feltner DE and Lalonde RL (2005) How modeling and simulation have enhanced decision making in new drug development. Journal of Pharmacokinetic and Pharmacodynamic 32: 185-197 167. Food and Drug Administration (2004) Innovation or stagnatio n? Challenge and opportunity on the critical path to new medical products. 127

PAGE 128

128 BIOGRAPHICAL SKETCH Navin Goyal was born in 1980 in Pune, India. He completed his Bachelors in Pharmacy from Maharashtra Institute of Pharmacy in Pu ne in 2002. He worked for Torrent Pharmaceuticals for a year. He then joined Nicholas Piramal India Ltd, another pharmaceu tical company where he worked for a year before joining the graduate program in pharmaceutics in August 2004 at University of Florida.