Citation
Factors Affecting Pulmonary Targeting of Inhaled Corticosteroids

Material Information

Title:
Factors Affecting Pulmonary Targeting of Inhaled Corticosteroids
Creator:
WU, KAI
Copyright Date:
2008

Subjects

Subjects / Keywords:
Asthma ( jstor )
Corticosteroids ( jstor )
Dosage ( jstor )
Inhalation ( jstor )
Lungs ( jstor )
Modeling ( jstor )
Pharmacokinetics ( jstor )
Plasmas ( jstor )
Propionates ( jstor )
Simulations ( jstor )
City of Gainesville ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Kai Wu. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
3/1/2007

Downloads

This item has the following downloads:


Full Text











FACTORS AFFECTING PULMONARY TARGETING OF INHALED
CORTICO STEROID S















By

KAI WU













A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


2006
































Copyright 2006

by

Kai Wu



































For my Waipo, Baba, Mama, Gege and Laopo.
















ACKNOWLEDGMENTS

I extend my appreciation and grateful thanks to Dr. Guenther Hochhaus for guiding

me through the four years of Ph.D. study, giving me excellent suggestions and always

supporting me. I would like to thank the members of my supervisory committee, Dr.

Hartmut Derendorf, Dr. Jeffrey Hughes, and Dr. Saeed R. Khan, for their valuable and

kind advice throughout my research. I take this opportunity to express my gratitude to

Yufei Tang for her invaluable technical assistance.

I would like to thank all the secretaries, Jame Ketcham, Andrea Tucker and Patricia

Khan, for their technical and administrative assistance. I thank my group members,

Yaning, Manish, Intira, Vikram, Sriks, Zia, Elanor, Keerti, Navin, Nasha and other

graduate students and post-docs in the department for their assistance and friendship.

Especially I want to thank all my friends, Weihui, Chao, Jeff, Victor, Aixin, Kai and

Zhen, Chengguan and Qin, Yipeng and Hongxiao and all my other friends for their help

and support that have made my stay in Gainesville a pleasant experience.

I want to express my deep gratitude and love for my parents, brother and other

family members for their unconditional support and encouragement of my academic

pursuits over the years.

Last, but not least, I want to thank my lovely wife, Ling, for her 120% support, full

understanding and timely encouragement. Without her, I would not have been able to

achieve any of what I have achieved.





















TABLE OF CONTENTS


page

ACKNOWLEDGMENT S ............ ..... __ ............. iv....


LI ST OF T ABLE S ............ ..... __ ............. viii..


LI ST OF FIGURE S .............. .................... ix


AB STRAC T ................ .............. xi


CHAPTER


1 INTRODUCTION ................. ...............1.......... ......


Asthma ................... ............... ......... .. ........ .........
Mechanism of Action of Corticosteroids in Asthma ................ ............... ...._...6

Disposition of Inhaled Corticosteroids ..........._..__....._.. ....._._ ...........8
Inhalation Device...................... ..............

Physical Characteristics of Patient..........._.._.. ... ............ ........__. ........ 1
Pharmacokinetic and Pharmacodynamic Properties of ICS ................ ........_.._... ...14
Pharmacokinetic Properties ................. ....._.._......_. ...........1
Prodrug ........._._ ......_ _. ............... 14...
M ucociliary Clearance .............. ...............15....
Pulmonary Residence Time .............. ...............15....
Bioavailability .............. ...............16....
Clearance ................. ...............17.................

Drug Distribution .............. ...............18....
H alf-life ................. ................. 19..............
Protein Binding .............. ...............19....
Obj ectives of the Study ................. ...............21........... ...

2 IN-VITRO COMPARISON OF FLUTICASONE RESPIRABLE DOSE FROM
A METERED DOSE INHALER AND THREE RIGID VALVED HOLDING
CHAM BERS .............. ...............22....


Introducti on ................. ...............22.................
M materials and M ethods .............. ...............23....
Chemical s and Devices............... ...............23

Sample Collection .............. ...............23....
Fluticasone Propionate As say............... ...............26.












Data Analysis............... ...............26
R e sults........._... .... ....... ...............27.....
Discussion and Conclusion............... ...............2


3 PLASMA CONCENTRATIONS AND PHARMACOKINETICS PROFILES OF
INHALED CORTICOSTEROIDS AND THEIR RELATION TO LUNG
FUNCTION IN ASTHMA ................. ...............32........... ....


Introducti on ................. ...............32.................
M ethods .............. ...............33....

Subj ects ................. ...............33.................
M easurements ................. ...............33.................

Lung function ................ ...............33.................
Plasma drug assays............... ...............33.
Protocol............... ...............34

Analy si s ................. ...............3.. 5..............
R results .............. ..... ........ ... .... .......3
Plasma drug concentrations and pharmacokinetic profiles ..........................37
Relation between lung function and AUC .............. ....................3
Discussion ................. ............. .. .......... ..... .... .... .......4

Plasma Drug Concentrations and Pharmacokinetic Profiles ............... ...............41
Relation Between Lung Function and AUC ................................... 4
Conclusion ................ ...............43.................


4 FLUTICASONE AND BUDESONIDE CONCENTRATIONS AFTER
INHALATION EFFECT OF BRONCHOCONSTRICTION ................. ................45


Introducti on ................. ...............45.................
M ethods .............. ...............46....

Subj ects ............. ...... ...............46....
M easurements ............. ...... ._ ...............46....

Spirometry .......___............ ....... ........___.........46
Methacholine inhalation challenge ........._...__ ......... ......._..._.......46
Drug assays .............. ...............47....
Protocol............... ...............47

Analy si s ............_. ...._... ...............48....
Re sults............._. ...._... ...............48....
Discussion ............._. ...._... ...............51....


5 EVALUATION OF THE ADMINISTRATION TIME EFFECT ON THE
CUMULATIVE CORTISOL SUPPRESSION AND CUMULATIVE
LYMPHOCYTES SUPPRESSION FOR ONCE-DAILY INHALED
CORTICO STEROID S............. ..... ._ ...............55.


Introducti on ............. ..... ...............55...
M ethods .............. ...............57....
Historic Data............... ...............57..












Population Pharmacokinetic/Pharmacodynamic Analysis ................. ...............58
Pharmacokinetic analysis .............. .................. ...............5
Pharmacokinetic/Pharmacodynamic analysis of effects of BUD and FP
on cortisol .............. ............. .... ..... .......5
Pharmacokinetic/Pharmacodynamic analysis of effects of budesonide
and fluticasone propionate on lymphocytes ............... ........ ...............6
Population PK/PD Simulation to Evaluate Administration Time Effect on
CCS and CLS for Once-daily Dose of BUD and FP..........._.. ..........._._.62
Re sults........._........ ...._.. ........._...... ...........6
Population Pharmacokinetic Model .............. ...............63....
Budesonide ................. ...............63.................
Fluticasone propionate .............. .. .........................6
Population Pharmacokinetic/Pharmacodynamic Model ............ .........._......65
Population Pharmacokinetic/Pharmacodynamic Simulation .............. .............66
Discussion ........._.... ...............73.____........
Conclusion ........._.... ...............78.____........


6 CONCLUSIONS .............. ...............80....

APPENDIX


A NONMEM CODE FOR PK MODEL OF BUD AFTER INHALATION .................83


B NONMEM CODE FOR PK MODEL OF FP AFTER INHALATION ................... ..84


C NONMEM CODE FOR PK/PD MODEL OF BUD .............. .....................8

D NONMEM CODE FOR PK/PD MODEL OF FP .............. ...............92....


E NONMEM CODE FOR BUD PK/PD SIMULATION MODEL. .............. ..... ..........98


F NONMEM CODE FOR FP PK/PD SIMULATION MODEL. .............. .............103

LIST OF REFERENCES ............ ...... ..__ ...............108..

BIOGRAPHICAL SKETCH ............_...... .__ ...............122...


















LIST OF TABLES


Table pg

3-1 Baseline characteristics .............. ...............36....

3-2 Mean (SD) pharmacokinetic parameters for beclometasone
monopropionate(BMP),budesonide(BUD), fluticasone propionate(FP) and
mometasone furoate (MF) ................. ...............38................

3-3 Partial correlation coefficients (r) for the relationship between FEV1 %
predicated and AUC after adjusting for age or gender............... .................4

3-4 Predicted AUC values at 50 and 100% predicted FEV1 and PEF and the ratio of
these values .............. ...............40....

4-1 Baseline characteristics .............. ...............49....

4-2 Pharmacokinetic parameters for fluticasone and budesonide when inhaled with
and without prior methacholine-induced bronchoconstriction............... ..........5

4-3 Area under the curve (AUC) values for fluticasone and budesonide when
inhaled with and without prior methacholine-induced bronchoconstriction............52

5-1 Summary of BUD pharmacokinetic model parameter estimates ................... ..........65

5-2 Summary of FP pharmacokinetic model parameter estimates ................ ...............67

5-3 Summary of FP and BUD pharmac oki neti c/pharm acodynami c m odel p aram eter
estimates ................ ...............70.................

5-4 Summary of the mean cumulative cortisol suppression of the 75 subjects of the 8
sim ulations .............. ...............73....

5-5 Summary of the mean cumulative lymphocytes suppression of the 75 subj ects of
the 8 simulations ................. ...............73........... ...

















LIST OF FIGURES


Figure pg

1-1 Molecular structures of several inhaled corticosteroids. .............. .....................

1-2 Cellular interactions after antigen inhalation in asthma ................. ............. .......4

1-3 Classical model of glucocorticoid (GCS) action. .........._._ ........... ...............8

1-4 Disposition of ICSs after inhalation ...._.._.._ ........__. ...._.._ ...........1

2-1 Schematic diagram of the cascade impactor with aerosol aerodynamic size range
at each impactor stage 0-7 and at the final filter............... ...............24.

2-2 Mean & SD respirable dose of fluticason propionate from metered-dose inhaler
(MDI) and valved holding chamber (VHC) devices. ............... ...................2

2-3 Mean & SD oropharyngeal dose of fluticason propionate from metered-dose
inhaler (MDI) and valved holding chamber (VHC) devices ................. ................29

3-1 Mean (SE) plasma concentration-time curves for beclometasone
monopropionate (BMP), budesonide (BUD), fluticasone propionate (FP) and
mometasone furoate (MF) following inhalation .............. ...............37....

3-2 Relation between FEV1 % predicted and the area under the plasma
concentration-time curve (AUC) for beclometasone monopropionate (BMP),
budesonide (BUD), fluticasone propionate (FP) and mometasone furoate (MF)
following inhalation .............. ...............39....

4-1 Mean (SE) plasma drug concentrations following inhalation of 1000 Cpg
fluticasone (Accuhaler@) and 800 Cpg budesonide (Turbohaler@) in 20 subj ects
with asthma with and without prior methacholine-induced bronchoconstriction ....51

5-1 Relationship between population predicted (open triangle) and observed BUD
(open circle) concentrations, individual predicted (open square) and observed
BUD (open circle) concentrations as function of time ................. ............. .......64

5-2 Relationship between population predicted (open triangle) and observed FP
(open circle) concentrations, individual predicted (open square) and observed FP
(open circle) concentrations as function of time. ............. ........ .............6











5-3 Diagnostic plots of BUD PK/PD model ................. ...............68........... ..

5-4 Diagnostic plots of FP PK/PD model ................. ...............69........... ..

5-5 Relationship between administration time and corresponding mean (+ SE)
cumulative suppression .............. ...............71....

5-6 Relationship between administration time and corresponding mean (+ SE)
cumulative suppression .............. ...............72....
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

FACTORS AFFECTING PULMONARY TARGETING OF INHALED
CORTICO STEROID S

By

Kai Wu

December 2006

Chair: Guenther Hocchaus
Major: Pharmaceutical Sciences

Understanding the factor affecting pulmonary targeting would help us improve

inhaled corticosteroids (ICSs) therapy. The three factors important for pulmonary

selectivity are inhalation device, characteristics of patients, and

pharmacokinetic/pharmacodynamic properties of ICSs. Certain aspects related to these

factors were evaluated in the current dissertation.

The effect of different valved holding chambers on in vitro aerosol deposition from

a fluticasone MDI was assessed using an in vitro cascade impactor method. It was found

that the mean fluticasone respirable dose from AeroChamber-PlusTM and OptiChambero

was no different (p > 0.05) from the respirable dose produced by the MDI alone, while

the OptiChambero'-Advantage respirable dose was significantly less compared with the

dose from the MDI alone (p < 0.05).

The overall goal of the second and third study was to test the hypothesis that the

pulmonary systemic availability of ICSs is related to lung function of the patients while









the extent of lung function affecting the bioavailability depends on the physicochemical

properties of ICSs. In the second study, it was found that the bioavailability for all four

drugs (beclomethasone, budesonide, mometasone and fluticasone) tended to be higher in

the asthmatic patients with higher lung function, while the associations between the

bioavailability and lung function differed for the four drugs. In the third study, it was

found that AUCO-5hour ValUeS for fluticasone and budesonide were lower by a median of

60% and 29% respectively when bronchoconstriction was induced in the asthmatic

pati ents.

Finally, the effect of administration time on cumulative cortisol (CCS) and

cumulative lymphocytes suppression (CLS) was evaluated for once-daily ICS and a

population pharmacokinetic/pharmacodynamic modeling/simulation approach was used.

It was found that the optimal time for administration of ICS depends on the choice of the

biomarker: afternoon in terms of minimized CCS, and morning in terms of minimized

CLS.















CHAPTER 1
INTTRODUCTION

More than 7% of adults and 12% of children in the United States (US) are affected

by asthma, a chronic inflammatory disease of the airways [1]. Over the world, asthma

afflicts 100 to 150 million people. The prevalence of asthma in the US has been

increasing by 5% to 6% over the past decades [2]. Asthma is responsible for 100 million

days of restricted activity and 470,000 hospitalizations each year [1]. In 1998, asthma

related cost was estimated at $11.3 billion in the U. S. alone.

Inhaled corticosteroids (ICSs) have revolutionized the treatment of asthma in the

past 30 years and now represent the first-line therapy for patients with chronic disease [3].

Numerous clinical studies have demonstrated the high anti-asthmatic efficacy (e.g.

prevention and control of symptoms) and reduced systemic side effects of ICS therapy.

Introduction of beclometasone dipropionate in 1972 as a pressurized metered-dose

inhaler (MDI) marked the beginning of the targeting treatment for asthma [4]. Since then,

other potent topical corticosteroids have been developed, and currently flunisolide (FLU),

triamcinolone acetonide (TA), budesonide (BUD), fluticasone propionate (FP),

mometasone furoate along with BDP (Figure 1-1) are available in the US. Additionally,

ciclesonide (Figure 1-1) is another ICS currently under development for the treatment of

asthma of all severities.

The use of inhaled corticosteroids for asthma allows, in effect, local delivery of the

drug to the lung. The goal of inhaled corticosteroids is to produce long-lasting therapeutic

effects at the pulmonary target site and to minimize systemic effects (pulmonary











selectivity). In general, pulmonary targeting of ICS is determined by their


pharmacokinetic and pharmacodynamic (PK/PD) properties, the biopharmaceutical


aspects of the delivery device and the characteristics of the patient. In this chapter, an

overview of factors affecting pulmonary targeting of ICS will be discussed. In addition,


the current literature on asthma pathogenesis and the mechanism of action of ICS in


asthma therapy is reviewed.




OH
noI S-CH- F,
C- O--C -C,H,





Beecometthaeone dipropionate Fluticasone propionate

~H-OH "'

o-c OG~Ho-





Triamcinolone ac~etonide Mornetasone furoate

O /CH,





CH-OH ~"--







Flunisolide ilsnd







Figure 1-1. Molecular structures of several inhaled corticosteroids [5, 6].









Asthma

Asthma is a chronic disorder of the airways characterized by airway inflammation,

reversible airflow obstruction and airway hyperresponsiveness. Studies have shown that

airway inflammation is an integral element in the pathogenesis of asthma and a driving

force in airway hyperresponsiveness and a contributing factor for variable levels of

airflow obstruction [7].

Asthma inflammation involves many cells, with mast cells, eosinophils, basophils,

neutrophils, macrophages, epithelial cells, and lymphocytes all being active participants

(Figure 1-2). The interactions of these cells result in the release of different mediators

leading to bronchoconstriction, microvascular leakage, mucus hypersecretion, epithelial

damage, and stimulation of neural reflex. For example, activation of the mast cell by

specific antigen through cell-bound IgE releases histamine and causes synthesis of

leukotrienes, which can precipitate an acute episode of airway obstruction [8]. Mast cells

also release inflammatory enzymes or proteases, which can contribute to airway

inflammation and airway hyperresponsiveness. Finally, mast cells express and release a

wide variety of proinflammatory cytokines including granulocyte-macrophage colony-

stimulating factor (GM-CSF), interleukin (IL)-3, 4 and 5 [9]. These cytokines change the

immunologic environment of the airway and are likely instrumental in initiating the

inflammatory response in asthma through changes in adhesion proteins, cell recruitment,

and inflammatory cell survival. Eosinophils are recruited to the lung in asthma after

antigen inhalation and are the main effector cells in the resulting inflammatory process.

Eosinophils contribute to inflammation by release of mediators (i.e., leukotrienes and

toxic oxygen products), by degranulation and release of granular proteins, and by

generation of cytokines [8, 10]. Granular proteins, such as maj or basic protein, directly










damage airway tissue, promote airway responsiveness, and injure airway epithelium. In

general, the degree of airway cosinophilia is usually proportional to the severity of

asthma [1l]. Activated lymphocytes are another important element of airway

inflammation. One of the T helper (TH) lymphocytes, TH2, Shows a cytokine profile that

contains IL-4 and IL-5, two cytokines that are essential for IgE synthesis (IL-4) and

eosinophil production (IL-5). Since lymphocytes are long lived, they are believed to be

crucial to the persistence of the inflammatory process in asthma. Epithelial cells are

another source of proinflammatory mediators in asthma including eicosanoids, cytokines,

chemokines, and nitric oxide [12].




Antigen



T-lymharolphagyer CeNeutrophil




Globet cell E~pithelial shedding



Bronchocoxnstriction


Figure 1-2. Cellular interactions after antigen inhalation in asthma [13]


The current concept of asthma therapy is based on a stepwise approach, depending

on disease severity, and the aim is to reduce the symptoms that result from airway

obstruction and inflammation, to prevent exacerbations and to maintain normal lung

function [14]. Controlling the allergic response to the local environment due to smoking

and allergens found at home or in the workplace, is the first priority in minimizing









asthma hyperresponsiveness. Traditionally, there are five classes of medications for

asthma therapy: corticosteroids (e.g., beclomethasone dipropionate), short-acting (e.g.,

albuterol) and long-acting (e.g., salmeterol) beta-agonists, theophylline, cromolyn sodium

and nedocromil sodium, and anticholinergics (e.g., ipratropium bromide) [15].

Corticosteroids, cromolyn sodium and nedocromil sodium can be classified as anti-

inflammatory medications. Beta-agonists and theophylline are both brochodilators.

Anticholinergics work by blocking the contraction of the underlying smooth muscle of

the bronchi. Among these medications, inhaled corticosteroids are the most effective for

the treatment of asthma. Their efficacy is related to many factors including a diminution

in inflammatory cell function and activation, stabilization of vascular leakage, a decrease

in mucus production, and an increase in beta-adrenergic response. The strong scientific

rationale behind the combination therapy of long-acting beta-aganists and corticosteroids

has led to the development of two mixed combination inhalers: salmeterol/fluticasone

and formoterol/budesonide. They are increasingly used as a convenient controller in

patients with persistent asthma. Advances in the knowledge of the pathophysiology of

asthma in the past decade have opened the way to the development of novel therapeutic

strategies that target more directly on the pathophysiology of asthma than those currently

in use [16]. Various compounds able to interfere with the complex network of

proinflammatory mediators, cytokines, chemokines, and adhesion molecules involved in

the pathogenesis of asthma have been identified, such as leukotriene modifiers,

tachykinin antagonists, endothelin antagonists, adenosine receptor inhibitors, tryptase

inhibitors, cytokine and chemokine inhibitors, direct inhibitors of T-cell function and

adhesion molecule blockers. New forms of immunotherapy, such as anti-IgE therapies,










gene vaccination with plasmid DNA and mycobacterial preparations, are also emerging

to aim at blocking the unbalanced TH2 TOSponse that characterizes the pathophysiology of

asthma. Despite the significant expansion of the experimentally available treatment

options, only the leukotriene modifiers are on the market, among which leukotriene

antagonists are now most frequently used in asthma treatment [17, 18]. The leukotriene

antagonists have been shown to inhibit exercise-provoked bronchospasm, to modify the

airway response to inhaled antigen, and to improve airway function in patients with

chronic asthma. But in head-to-head trials with inhaled corticosteroids, the leukotriene

antagonists are less effective in terms of improvement in lung function and reduction in

exacerbations [7]. Thus far, inhaled corticosteroids still represent the cornerstone of

treatment for asthma.

Mechanism of Action of Corticosteroids in Asthma

Corticosteroids are able to affect many of the inflammatory pathways involved in

the pathogenesis of asthma, including the complex cell to cell communications mediated

by the so-called "cytokine network" [19]. Corticosteroids start their action when the

lipophilic corticosteroid molecule crosses the cell membrane and binds to a specific,

intracellular glucocorticoid receptor (GR) (Figure 1-3). The GR is located in the

cytoplasma and belongs to the steroid thyroid/retinoic acid receptor superfamily [20].

Prior to binding corticosteroids, GR forms a large heteromeric complex with several

other proteins, from which it dissociates upon ligand binding. 90 kDa heat shock protein

(hsp90) plays a central role in this complex [21]. It binds as a dimer to the C-terminal

ligand binding domain (Figure 1-3). The association with hsp90 has been shown to be

required to maintain the C-terminal domain of GR in a favorable conformation for ligand

binding [22], but GR is transcriptionally inactive at this stage. Binding of corticosteroids









to GR induces dissociating hsp90 and other heat shock proteins, resulting in a

conformational change which allows the GR complex to translocate to the nucleus or

interact with cytoplasmic transcription factors. Once in the nucleus, the GR complex

binds as a dimer to specific DNA sites, a specific nucleotide palindromic sequences

termed "glucocorticoid response elements" (GRE) (Figure 1-3) [23]. Then the

transcription of specific genes can be increased (transactivation) or decreased

(transrepression) depending on whether the GRE is positive or negative. In particular,

corticosteroids increase the expression of anti-inflammatory proteins such as lipocortin-1,

interleukin-1 receptor antagonist (IL-1ra), interleukin-10 (IL-10), secretary leukocyte

inhibitory protein, neutral endopetidase and the inhibitory protein (IkB) of nuclear factor-

KB (NF-KB) [3]. It is known that lipocortin-1 inactivates the enzyme phospholipase A2

thus inhibiting the production of lipid mediators (i.e., platelet-activating factor,

leukotrienes, and prostaglandins) and NF-KB positively regulated inflammatory and

immune responses [24, 25]. Corticosteroids are also able to enhance the transcription of

the gene encoding the P2-adrenergic receptor, thus reversing and/or preventing the down-

regulation possibly induced by long-term treatments with P2-agOnist bronchodilators [13].

However, the genomic mechanism of transactivation is believed to be mainly responsible

for the unwanted side effects of corticosteroids, rather than for their anti-inflammatory

and immunoregulatory actions. The very effective control of airway inflammation

exerted by corticosteroids in asthma is largely mediated by inhibition of the

transcriptional activity of several different genes encoding pro-inflammatory proteins

such as cytokines (IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-11, IL-13, TNF-1a2, GM-CSF),

chemokines (IL-8, RANTES, MIP-la, MCP-1, MCP-3, MCP-4, eotaxin), adhesion










molecules (ICAM-1, VCAM-1, E-selectin), and mediator-synthesizing enzymes (i-NOS,

COX-2, cytoplasmic PLA2), which are mostly due to GR-dependent repression of pro-

inflammatory genes [19]. Corticosteroids also exert their repressive action by direct

interaction with transcription factors and post-transcriptional regulation [19].




Ceti menthane
SLlpeosedn-1
|sTB~,drncepphersi Cylopinsm

tCytakinee
1INOS
mRanara\/1 COX-2






Nadeeus +GRE nGiRE Sto~drtr-rsponstre+
targer gene
Figure 1-3. Classical model of glucocorticoid (GCS) action. The glucocorticoid enters the
cell and binds to a cytoplasmic glucocorticoid receptor (GR) that is
completed with two molecules of a 90kDa heat shock protein (hsp 90). GR
translocates to the nucleus where, as a dimer, it binds to a glucocorticoid
recognition sequence (GRE) on the 5-upstream promoter sequence of steroid-
responsive genes. GREs may increase transcription and nGREs may decrease
transcription, resulting in increased or decreased messenger RNA (mRNA)
and protein synthesis (Adapted from Barnes [26]).


Disposition of Inhaled Corticosteroids

Understanding the disposition of ICS after inhaled administration will help us

understand why PK/PD properties, the choice of delivery device and the physical

characteristics of the patient are important for achieving pulmonary targeting of ICS


(Figure 1-4).










Upon inhalation of a glucocorticoid, only 10% to 60% of the administered dose is

deposited in the lungs. The other of the dose (40% to 90%) impacts on the oropharyngeal

region and is swallowed. The swallowed dose is absorbed from the gastrointestinal tract

and after undergoing first pass inactivation by the liver only orally bioavailable fraction

enters the systemic circulation. The drug particles deposited in the lungs dissolve

according to the rate of dissolution of the drug. If not readily dissolved, the drug will be

removed by mucociliary clearance back into the oropharyngeal region and then is

swallowed. Only the dissolved drug interacts with the glucocorticoid receptors (GR), as

described in the previous section, present in the lung to produce the anti-inflammatory

effects. Following the interaction with the GR, the drug is absorbed into the systemic

circulation. The absorption processes are rapid and it is believed that dissolution is the

rate-limiting step involved in pulmonary absorption of IGCs. 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 GR to produce the systemic side effects. The drug is eventually eliminated

from the systemic circulation mainly by hepatic clearance mechanisms.

Inhalation Device

One factor to be considered for distinct pulmonary selectivity is efficient delivery

to the lungs. As mentioned earlier, about 10% to 60% of nominal inhaled dose will

deposit in the lung. The exact percentage and the distribution of the dose in the lung

depends great deal on the biopharmaceutical aspects of the delivery device. It also

depends on the patient since every inhaler requires certain level of technique for optimal

delivery .









There are three main classes of inhalation devices for ICSs: metered-dose inhaler

(MDI), breath-activated dry powder inhaler (DPI) and nebulizer. Inside MDI the drug

exists as a suspension in a carrier liquid or a solution delivered through a

chlorofluorocarbon (CFC) or hydrofluoroalkane (HFA) propellant, respectively [27].

However, CFC-MDI' s are gradually phased out because of their ozone-depleting

potential. In addition to their environmentally friendly property, HFA solutions also seem

to have the advantage of delivering a much greater mass of fine particles, with a diameter

of less than 5 lm. Fine particles are more likely to be deposited in the tracheo-bronchial

and pulmonary regions in the lung. On the other side, larger particles are deposited

mostly in the oropharynx where they are swallowed and increase the risk of systemic

absorption [28]. The average particle diameter delivered by a CFC-MDI is 3.5-4.0 Clm

whereas the average particle diameter delivered by a HFA propellant is around 1.1 lm.

This difference in particle diameter might have a clinical significance as the average

diameter of small airways is around 2Clm, resulting in a greater lung deposition [29]. This

increased proportion of fine particles with the HFA-MDI results in an improved lung

deposition. A previous study showed the lung deposition of beclometasone dipropionate

(BDP) was increased from 4-7% using CFC-MDI to 55-60% using a newly developed

HFA-MDI [30]. Another study showed with the CFC-free solution MDI, a mean lung

deposition of 52% for ciclesonide could be obtained [31]. In a single-dose study

comparing HFA flunisolide and CFC flunisolide, the drug deposition in the lung could

even be increased to 68% (HFA) compared to 19.7% (CFC) [32]. Hence, replacing the

propellant CFC with HFA leads to an increase in lung deposition. Lung deposition can

also be increased by use of spacer and valved holding chamber(VHC) devices, which can


























:-~ Mouth and Pharynx


Mucociliary clearance


alter the amount of fine particles and therefore, increase the respirable fraction and

decrease the amount of drug deposited in the oropharynx, [33]. Moreover, use of VHC

with MDI helps the patients who lack coordination between actuation and inhalation.

However, it also needs to be kept in mind that a greater lung deposition might result in a

greater possibility of systemic adverse effects because of the lack of first-pass

metabolism after direct absorption from the lung.


Figure 1-4. Disposition of ICSs after inhalation (Adapted from [34])


The other inhaler type used for ICSs is the dry powder inhaler. Use of DPI requires

easier delivering technique that requires less coordination than the MDI. However, it

requires a forceful deep inhalation to trigger the inhalation device to help break up the

aggregates of the micronized powders into respirable particles in the oropharynx and

larger airways. Therefore, lung deposition is flow-dependent and the higher the inhalation

flow, the smaller the particles will be [35]. An inspiratory flow of 60L/min is considered


Orally absorbed I~laac
Fraction Clearance

System ic
,ass Side Effects
ation


Fi rst-p
Inactiv;









to be optimal [33]. Hence, it should be ensured that asthmatic patients in all asthma

stages are able to achieve an inhalation flow that is enough to achieve the required effect

[35]. In a lung deposition study it was shown that reducing the inhalation flow from 58

L/min to 36 L/min reduces the lung deposition of budesonide from around 28% to around

15% [36].

The standard and most common used j et nebulizer is based on a constant output

design, run by compressed air or oxygen with supplemental air drawn in across the top of

the nebulizer. The nebulizer contains the drug either as a suspension or a solution. There

have been also new developments in the field of nebulizers and liquid formulations.

Among those are the inhalation device MysticTM fTOm Batelle, which is based on electro-

hydrodynamic principles, employing electrostatic energy to create fine aerosols from

formulated drug solutions or suspensions thereby increasing the pulmonary tract

deposition to about 80% [37] and the RESPIMAT device from Boehringer Ingelheim

uses a high-pressure micro-spray system of nozzles to release a metered dose to the

patient. This system generates a slow release of the drug with a high concentration of

respirable particles[38].

Improved pulmonary deposition will lead to improved pulmonary targeting,

especially for drugs with significant oral bioavailability such as BDP, as efficient

pulmonary delivery will reduce the amount of swallowed drug and, subsequently, oral

absorption. In addition, the new more efficient pulmonary delivery devices reduce oral

deposition which may also reduce oropharyngeal adverse effects.[39] However, efficient

pulmonary deposition, with respect to systemic adverse events, is not as important for










drugs with low oral bioavailability, such as ciclesonide, mometasone furoate and

fluticasone propionate, as very little drug will be absorbed through the oral route anyway.

Physical Characteristics of Patient

Many aspects of physical characteristics of the patient could affect pulmonary

targeting of a ICS therapy. One of them is age. Usually, a proper technique for use of

inhalation device is much easier to be learned and employed among adults. On the

contrary, children and elderly might not learn to use the device properly, which could

leads to less or more dose than right dose taken and reduced pulmonary selectivity of ICS

therapy. The other issue related with age is the compliance to dosing regimen. In a study

evaluating compliance with ICSs, it was found that the children between 8 and 12 years

only averaged 58% of compliance [40]. Furthermore, only 32% of those doses were

actually received at the correct time. A non-compliance with dose is highly likely to

associated less efficacy of the therapy. Moreover, non-compliance with timing of the

dose would probably affect the pulmonary selectivity of therapy, as it has been shown by

some other studies that administration time is an important factor to maximize the

efficacy and minimize the systemic side effects of ICSs [41-48] The important of

administration time on systemic side effects of ICSs will be discussed in later chapter of

this dissertation.

Another important aspect of patient characteristics affecting pulmonary selectivity

of ICSs is the lung function. It has been suggested that the pulmonary systemic

bioavailability of certain ICSs is associated with the lung function of patient [49]. In one

study with a single dose fluticasone dry powder in patients with asthma of varying

degrees of severity there was a highly significant linear correlation between the absolute

magnitude of adrenal suppression and the lung function expressed as percentage









predicted forced expiratory volume in 1 second (FEV1) [50]. This is consistent with

pharmacokinetic data where there was 62% lower plasma fluticasone concentrations (as

AUC) in patients with moderately severe asthma (FEV1 = 54% predicted) than in healthy

volunteers receiving inhaled fluticasone. However, in another study, there was no

difference in plasma budesonide concentrations between healthy volunteers and patients

with mild asthma after inhalation budesonide, a much less lipophilic compound than

fluticasone [51]. The correlation between lung bioavailability, lung function and drug's

lipophilicity will be discussed in later chapters.

Pharmacokinetic and Pharmacodynamic Properties of ICS

Pharmacokinetic Properties

Prodrug

A prodrug is a pharmacologically inactive compound that is activated in the body

after its administration. To exert a local effect, a prodrug needs to be activated in the

target tissue, e.g. lung. As an example, the ester functional group in ciclesonide needs to

be cleaved by esterases in the lungs, before the active compound desisobutyryl-

ciclesonide can interact with the receptor [6, 52]. The use of a prodrug might be

beneficial as high lipophilicity may increase the pulmonary residence time. However,

activation predominantly in the lungs is required to maintain pulmonary selectivity. One

additional advantage of drugs that are delivered in an inactive form is that they might

reduce certain oropharyngeal adverse effects such as oral candidiasis. It has been shown

that the negligible activation of ciclesonide in the oropharyngeal region reduce the

oropharyngeal adverse effects [39, 53].









Mucociliary Clearance

The mucociliary transporter is able to remove those slowly dissolving solid drug

particles deposited to the upper part of the respiration tract. This phenomenon will be less

pronounced for rapidly dissolving ICS or those present in solution. Mucociliary clearance

mechanism is more pronounced in the central airways of the lung compared to the

peripheral airways and plays an important role in reducing the overall availability

(systemic + pulmonary) of highly lipophilic glucocorticoids that have a slow rate of

dissolution. Studies have shown that the systemic availability of more slowly dissolving

ICS, such as fluticasone propionate, is lower in patients with asthma when compared with

healthy volunteers [54, 55]. The reason is that deposition in patients with asthma tends to

be more central in the lungs.

Pulmonary Residence Time

The large surface area, thin membranes and the existence of pores allow an

efficient exchange between the outer part of the lung and the circulatory system. Once

dissolved in the lung lining fluid, ICS, as lipophilic drugs, will easily cross cell

membranes and be absorbed quickly and efficiently into the systemic circulation [56].

Therefore, to achieve pulmonary selectivity requires the drug efficiently delivered to the

lung, however a fast absorption into the systemic circulation immediately after delivery

shall be prevented. Triamcinolone acetonide, when given as solution into rat lungs, was

unable to produce pulmonary targeting as the receptor occupancy in lung and liver was

very similar [57]. Extending the residence time of the drug in the lung (e.g. through

controlling the release of the drug in the lung) will increase pulmonary selectivity as high

drug concentrations are present in the lung for a longer period of time [58]. Various

approaches to extend the pulmonary residence time of drugs have been evaluated in the









pre-clinical setting.[59, 60] To achieve increased pulmonary residence time, two

approaches are currently utilized for ICSs. Highly lipophilic drugs, such as fluticasone

propionate, will dissolve slowly after deposition in the lung as particles and insure high

lung concentrations over a longer period of time. A pharmacokinetic parameter used to

quantify pulmonary residence time is the mean pulmonary absorption time (the average

time the drug will need to leave the lung). Commercially available drugs differ in this

property, with fluticasone propionate showing the highest pulmonary residence time of

almost 5 hours while flunisolide is absorbed very rapidly. An alternative biological "slow

release system" has been identified for budesonide and potentially ciclesonide [61-63]. In

pulmonary cells, these two drugs form lipophilic esters that are biologically inactive but

are so lipophilic that they are unable to cross membranes therefore they are trapped inside

the cell to form a reservoir of active drug since the esters will be slowly cleaved back to

the active drug. This esterfication/ester-cleavage system, therefore, ensures a long

residence time in the lung which might provide sufficiently prolonged asthma control for

a reduced dosing regimen with budesonide and ciclesonide.

Bioavailability

The bioavailability of an inhaled corticosteroid is the rate and extent at which the

drug reaches its site of action (pulmonary bioavailability) as well as the blood (systemic

bioavailability). As mentioned earlier, a large fraction (approximately 40-90%) of the

inhaled dose is swallowed and subsequently available for systemic absorption. This

bioavailability of the orally delivered part is dependent on absorption characteristics of

the drug from the gastro-intestinal tract and the extent of intestinal and hepatic first-pass

metabolism. The oral bioavailabilities of currently used corticosteroids range from less

than 1% for fluticasone propionate to 26% for 17-beclomethasone monopropionate [64-









70]. However, the main determinant of systemic bioavailability after inhalation is direct

absorption from the lung, where for the currently available inhaled corticosteroids there is

no first-pass effect. All of the drug that is deposited in the lung will be absorbed

systemically [49]. The percentage of the drug dose that is deposited in the lung is greatly

influenced by the efficiency of the delivering device. Since the orally absorbed fraction of

the drug does not contribute to the beneficial effects but can induce systemic side effects,

it is desirable for the oral bioavailability of inhaled corticosteroids to be very low.

Therefore, the pulmonary bioavailability is rather a function of the delivery device used

for inhalation than a property of the drug itself. The pulmonary bioavailability will

depend on the amount deposited in the lungs and will differ with the delivery device used

[49, 71]. Fluticasone propionate, for example, has an oral bioavailability of <1% due to a

high first-pass metabolism (see above). When administered to the lungs using a dry

powder inhaler (DPI), the absolute bioavailability (systemic + pulmonary) is reported to

be approximately 17% and compared to 26% to 29% when using a metered dose inhaler

(MDI) [27, 72]. After mometasone fuorate administration via a dry powder inhaler the

absolute bioavailability was reported to be 11% [72].

Clearance

Clearance is a PK parameter describing the efficiency of the body to eliminate the

drug and quantifies the volume of the body fluid that is cleared of drug over a given time

period. For most drugs showing linear protein binding and non-saturated elimination this

value is a constant. The higher the clearance of an ICS the lower the systemic adverse

effects will be and the more pronounced the pulmonary selectivity. If drugs are mainly

metabolized and cleared by the liver, the maximum clearance possible is that of the liver

blood flow of ~90 L/hour. Several of the available ICS (budesonide, fluticasone










propionate) show clearance values close to the liver blood flow [73, 74]. The clearance of

such high extraction drugs is independent of protein binding. The active metabolites of

ciclesonide and beclomethasone propionate have been reported to express higher

clearance values when given intravenously [65, 75]. While these data suggests extra-

hepatic metabolism, future studies with the active metabolites need to be performed to

clearly confirm the high clearance of these two drugs, especially as this would represent a

clear advantage for these drugs over other commercially ICS.

Drug Distribution

The volume of distribution (Vd) is ai PK parameter that describes the distribution of

drug in the body. The higher the Vd, the greater the concentration of drug present in the

tissues and the lower the plasma concentration. For drugs such as ICS that can cross

membranes easily, the Vd is determined primarily by how much drug is bound to plasma

proteins and how much drug is bound to tissue components. Thus, corticosteroids with a

very large volume of distribution (300-900L) are extensively distributed and bound to the

tissues. The more drug that is bound in the tissue compartment (the smaller the fraction

unbound in tissue fut,) the higher will be the Vd. Similarly, a weak plasma protein binding

(fraction unbound drug in plasma fu is large) will allow more drug to enter the tissue

compartment. Because the Vd is determined by the ratio of protein binding in plasma and

tissue, it is possible for two drugs with significant differences in the plasma protein

binding to show the same Vd. This indicates that the Vd is a parameter that is difficult to

evaluate. Computer simulations did show that an increase in Vd, although affecting the

half-life of a drug, does not affect the pulmonary selectivity [76].









Half-life

Half-life (tl/2) is a secondary PK parameter that is determined by clearance and the

Vd [77]:

tl/2 0.693 *Vd/ClearTRCe

A short half-life, because of a high clearance, is beneficial for achieving pronounced

pulmonary selectivity. However, a short half-life due to a small Vd does not affect the

pulmonary selectivity [76]. The half-life is therefore of little value if an ICS has to be

evaluated with respect to its pulmonary selectivity.

Protein Binding

Protein binding is one of the important PK properties of ICS. The maj or protein

that is involved in binding with ICSs is albumin. Albumin comprises about 55-62% of the

protein present in plasma and also occurs plentifully in the tissues of mammals. Protein

binding is seen as storage form of the drug because it is reversible and a rapid equilibrium

is established. High plasma protein binding is considered as a desirable characteristic for

ICSs. Once ICSs reach the systemic circulation, binding to plasma proteins decreases the

free unbound fraction in the systemic circulation and thus decreases the risk for systemic

side effects, since only the free, unbound drug is pharmacologically active. Reported

clinical studies showed that ciclesonide, a novel inhaled corticosteroid with plasma

protein binding of 99% for itself and its active metabolite had an incidence of adverse

effects that was not significantly greater than that observed with placebo [31, 78, 79]. The

extent of plasma protein binding differs among ICSs and it has been positively associated

with their lipophilicity. Current ICSs in development tend to be relatively lipophilic.

Plasma protein binding has been reported to be 80% for flunisolide, 87% for

beclomethasone dipropionate, 88% for budesonide, 71% for triamcinolone acetonide,









90% for fluticasone propionate, 98% for mometasone furoate and 99% for active

principle of ciclesonide [79-82].

As discussed above, those ICSs with higher plasma protein binding would have a

safety margin over those ICSs with lower plasma protein binding. However, the increased

plasma protein binding may not necessarily improve the risk-benefit ratio. As lipophilic

compound, ICSs could penetrate cells and bind to both plasma and tissue proteins. It is

reasonable to expect the tissue protein binding for those ICSs with high plasma protein

binding would be high too because of their high lipophilicity. The fraction of free drug at

the target site would be reduced. The high tissue protein binding is also indicated by

ICSs' high volume of distribution. The steady state volume of distribution (Vss) of a drug

can be calculated according to:


Vss = Vp + x Vft]


where fup is the unbound fraction of drug in plasma, fut is the unbound fraction of drug in

tissue, Vp is the plasma volume and Vt is the volume of tissue compartment. The Vss of

ICSs could also be positively associated with their lipophilicity. Thus, for those relative

lipophilic IG with higher plasma protein binding and Vss, as we can see from above

equation, they would have increased tissue protein binding as well. An increase in tissue

protein binding of just a few percent can translate into a significant reduction in the

amount of efficacy. For example, if one was 99% bound in the lung and another was

about 98% bound in the lung, the one with 98% bound would present two folds of

unbound concentration for interaction with receptor to produce beneficial effects. A

higher dose may be needed for the ICSs with higher tissue binding to achieve efficacious

free concentration, thus the safety margin obtained by their high plasma protein binding









would be offset. Therefore, it is essential to determine both plasma protein binding and

tissue binding, especially the target tissue, lung protein binding for a complete evaluation

of efficacy and safety profile oflICSs.

Objectives of the Study

Pulmonary targeting is essential for successful inhaled corticosteroids (ICS). Many

factors determine the pulmonary targeting of ICS and these factors include

pharmacokinetic and pharmacodynamic (PK/PD) properties of ICS, inhalation device and

patients characteristics. Understanding of these factors has contributed significantly to

improve pulmonary targeting of modern ICS. This research has been focusing on

evaluation of some aspects related to these factors and the overall obj ectives could be

summarized as followings:

Aim 1: Compare the in-vitro respirable doses of fluticasone from a metered dose

inhaler and three rigid valved holding chambers.

Aim 2: Assess pharmacokinetic profiles of inhaled beclomethasone, budesonide,

fluticasone and mometasone in asthmatic subj ects with a broad range of lung function.

Aim 3: Compare the systemic exposure change between inhaled budesonide and

fluticasone after lung function decrease induced by methacholine in asthmatic subjects.

Aim 4: Evaluate administration time effect on the cortisol suppression and

lymphocytes suppression for once-daily inhaled corticosteroids.















CHAPTER 2
IN-VITRO COMPARISON OF FLUTICASONE RESPIRABLE DOSE FROM A
METERED DOSE INHALER AND THREE RIGID VALVED HOLDING
CHAMBERS

Introduction

There are several metered-dose inhaler (MDI) valved holding chambers (VHCs)

available in the United States. The primary purpose of VHC is to hold the MDI puff very

shortly before inhalation to minimize the detrimental effects of poor timing between MDI

actuation and patient inhalation. Another important feature of VHC is to allow the MDI

propellant time and distance to evaporate after actuation. Evaporation of propellant

promotes the formation of smaller aerosol particles that are likely to be carried into small

airways. Moreover, VHC collect "large" aerosol particles that would otherwise deposit

into the oropharynx. VHC has been recommended for all patients who can not effectively

use MDI alone, such as children and for all asthma patients prescribed an ICS to reduce

the risk of topical side effects of corticosteroid such as thrush and dysphonia [1].

In vitro aerosol data indicated that the VHC choice has no impact on the aerosol

characteristics of bronchodilators from an MDI [83]. Further, clinical studies have been

unable to discern measurable difference in bronchodilator effect among various albuterol

MDI-VHC combinations [84, 85]. Hence, many clinicians and third-party payers

conclude that all VHCs available in the USA markets are equivalent. However, this

statement may not stand for ICS as several in vitro studies indicate that the quantity of

ICS available for delivery to human lung can vary as many as three folds, depending on

the choice of VHC [83]. Therefore, interchanging one VHC with another may result in









pronounced fluctuations in the aerosol dose reaching the lungs and potentially result in

therapeutic failures. Fluticasone propionate MDI (Flovent, GlaxoSmithKline, Research

Triangle Park, NC) is the most often prescribed ICS in the USA for asthmatic patients.

Hence, to assess the aerosol characteristics of fluticasone propionate from various VHCs

is very important for delivery of successful therapy to asthmatic patients.

Materials and Methods

Chemicals and Devices

Acentone nitrile and trifluoroacetic acid were obtained from Fisher Scientific.

Beclomethasone dipropionate was kindly provided by 3M. Double distilled deionized

water was prepared in our lab (Gainesville, FL). The MDI tested was 110plg Flovento

(GlaxoSmithKline, Research Triangle Park, NC). The VHCs tested were AeroChamber-

PlusTM" (Monaghan Medical, Plattsburgh, NY), OptiChambero (Respironics, Murrysville,

PA) and OptiChambero'-Advantage (Respironics, Murrysville, PA). Andersen Mark-H,

eight-stage cascade impactor equipped with a USP induction port was purchased from

Thermo Andersen (Smyrna, GA).

Sample Collection

The deposition characteristics of fluticasone propionate aerosol delivered from the

MDI alone (110plg Flovento), the MDI attached to AeroChamber-PlusTM" VHC,

OptiChamberoVHC, and OptiChambero'-Advantage VHC were examined by using the

well-established in-vitro cascade impactor method [86, 87]. Fluticasone propionate

aerosol particles that entered the cascade impactor were sorted by aerodynamic diameter

onto eight different stages (Figure 2-1).































Figure 2-1. Schematic diagram of the cascade impactor with aerosol aerodynamic size
range at each impactor stage 0-7 and at the final filter (F). MDI = metered-
dose inhaler; VHC = valved holding chamber; USP = United States
Pharmacopeia.


Aerodynamic diameter does not refer to actual aerosol droplet diameter, rather it is

a means of categorizing particles that have different shapes and densities with a single

dimension based on how those particles behave when settling out. Aerosol deposition

onto the "throat" intake port was reflective of in vivo oropharyngeal deposition [88].

Aerosol deposition on stages 0, 1, and 2 of the cascade impactor corresponded to particles

of 5.1-10 lm. These particles are too large for therapeutic use in asthma and likely would

deposit into the large upper airways such as the trachea [89-91]. Aerosol that deposited

on stages 3, 4, and 5 of the cascade impactor corresponded to particles of 1-5 Clm, a

range considered ideal for deposition and retention within the small human airways

(bronchi and bronchioles) [89]. Aerosol collected on stages 6, 7, and in the final filter of

the cascade impactor corresponded to particles smaller than 1 Clm, a range considered too









small for therapeutic use in asthma and one that favors deposition into the respiratory

bronchioles and alveoli [89].

Fluticasone propionate aerosol from the MDI connected to each of six spacer

configurations and aerosol from the MDI alone were sampled directly into the cascade

impactor (Mark IIACFM eight-stage nonviable ambient sampler; Andersen Instruments

Inc., Atlanta, GA) through an artificial throat manufactured according to United States

Pharmacopeia (USP) standards [87]. Aerosol was drawn into the impactor by continuous

airflow generated by a vacuum pump (model 20-709; Graseby-Andersen, Atlanta, GA).

Airflow through the impactor was calibrated to conform to the manufacturer' s

requirements (28.3 L/min + 5%) and simulated a tidal inspiratory flow of about 30

L/minute. Before each sampling run, continuous airflow through the impactor was

allowed to equilibrate for 5 minutes, and each fluticasone propionate MDI canister was

primed 5 times in a separate room to ensure uniformity among actuations. Before each

sampling run, the MDI valve was washed twice with methanol and dried to remove any

trace drug that resulted from priming. Each MDI canister was shaken for at least 10

seconds before each actuation. Each of the above devices was tested Hyve times, and

during each run, 5 MDI puffs were sampled through the cascade impactor without delay

between actuation and sampling. After each actuation into the impactor, airflow was

maintained for 60 seconds before the next actuation. After the Einal actuation in each run,

flow through the impactor was maintained for an additional 60 seconds. Laboratory

personnel wore antistatic outer clothing containing carbon fibers during sample collection

to minimize transfer of electrostatic charges to the VHCs. All VHCs were washed in a

liquid detergent solution and allowed to air dry before use, to minimize electrostatic









attraction of fluticasone propionate aerosol during sample collection [92]. Ambient

temperature (22 & 20C) and humidity (46 & 4%) in the aerosol laboratory were monitored

to minimize inter-day variability in sample collection

Fluticasone Propionate Assay

After each trial, fluticasone propionate was washed from the USP throat and from

each of eight collection plates in the cascade impactor with methanol and an equal

volume of distilled water. The volume used to wash each section of the apparatus varied

from 5-20 ml to eliminate concentrations of fluticasone propionate below the assay limit

of detection (0.5 Clg/ml). Two 1-ml aliquots of the resultant solutions were collected for

high-performance liquid chromatography (HPLC) analysis. Fluticasone propionate

content from all washings was determined by inj ecting duplicate 100-C1l samples into a

reverse-phase HPLC system with a 254-nm ultraviolet detector. The HPLC mobile phase

consisted of acetonitrile and water (53:47 volume:volume ratio [v:v]) with 0.03%

trifluoroacetic acid. The flow rate through the system was 1.0 ml/minute. Fluticasone

propionate standard curves and quality controls were prepared in a methanol and water

mixture (50:50 v:v) Budesonide, as an internal standard was added to all solution before

analysis. Correlation coefficients for the standard curves were 0.998-1.000. Assay

interday coefficient of variation for fluticasone propionate quality control solutions

ranged from 11.4% for the 2.0-Clg/ml control to 3.2% for the 20.0-Clg/ml control.

Data Analysis

The amount of fluticasone propionate deposited on each stage and throat of the

cascade impactor was measured and normalized per 110 mg MDI actuation. A log

probability plot constructed from these data was used to determine the distribution of

particle sizes, i.e., mass median aerodynamic diameter (MMAD) and geometric standard









deviation (GSD) [87]. The quantity of fluticasone propionate aerosol emitted within the

ideal size range (1-5 Clm) for deposition and retention in the small human airways (i.e.,

respirable dose) and the quantity collected in the USP throat intake port (i.e.,

oropharyngeal dose) were also calculated. Differences among outcomes were determined

by using one-way analysis of variance. The Tukey multiple comparison procedure was

used to compare individual means when the overall test results were statistically

significant (p<0.05). All data analysis was performed with InStat statistical software,

version 3.01 (GraphPad Software Inc., San Diego, CA).

Results

MMAD of aerosols produced by the tested VHC devices were slightly smaller than

MMAD of aerosol produced by MDI alone (p < 0.05, Table 2-1). Similarly, the tested

devices reduced the GSD of aerosol compared to aerosol produced by the MDI alone (p <

0.05, Table 2-1). The differences in MMAD and GSD among the VHCs were not

significant (p > 0.05, Table 2-1).


Table 2-1: Mean (SD) MMAD and GSD values interpolated from the size distribution
data for MDI and VHCs.
MDI AeroChamber- OptiChambero OptiChambero-
PlusTM Advantage
MMAD 2.84 (0.32) 2.50 (0.05) 2.42 (0.14) 2.30 (0.06)
GSD 1.26 (0.04) 1.22 (0.01) 1.22 (0.01) 1.22 (0.01)

The fluticasone propionate respirable dose from all devices is shown in Figure 2-2.

The mean & SD fluticasone propionate respirable dose (i.e., aerosol size 1-5 mm) from

AeroChamber-PlusTM (45.3 A 1.8 Clg/ actuation) and old OptiChambero (38.1 & 8.7 Clg/

actuation) were not significantly different (p > 0.05) from the respirable dose produced

by MDI alone (46.7 & 9.7 Clg/ actuation). However, OptiChambero'-Advantage produced




























II


a mean respirable dose (32.1 + 1.6 Clg/ actuation) that was significantly less than that

produced by MDI alone or AeroChamber-PlusTM (p < 0.05) but not significantly different

from old OptiChambero (p > 0.05).


MDI alone
(46.7 & 9.7)


Aerochamber-Plus Optichamber
(45.31 1.8) (38.1 & 8.7)


Optichamber-Advantage
(32.1 1 1.6)


Figure 2-2. Mean & SD respirable dose of fluticasone propionate from metered-dose
inhaler (MDI) and valved holding chamber (VHC) devices.


The fluticasone propionate oropharyngeal dose from all devices is shown in Figure

2-3. The mean & SD oropharyngeal dose from MDI alone (50.5 & 7.2 Clg/ actuation) was

significantly greater (p < 0.001) than the oropharyngeal dose from all three VHC devices:

AeroChamber-PlusTM (0.9 & 0.3 Clg/ actuation), OptiChambero (0.6 & 0.8 Clg/ actuation)

and OptiChambero'-Advantage (0.17 & 0.1 Clg/ actuation). The difference in

oropharyngeal dose between the three VHCs was not significant (p > 0.05).


I


I













o











O MDI alone Aerochamber-Plus Optichamber Opti cham ber-Adva ntage
O ~(50.5 & 7.2) (0.9 & 0.3) (0.6 &0.8) (0.17 & 0.1)


Figure 2-3. Mean + SD oropharyngeal dose of fluticason propionate from metered-dose
inhaler (MDI) and valved holding chamber (VHC) devices.


Discussion and Conclusion

All three VHCs tested significantly decreased the oropharyngeal dose compared

with MDI alone. This reduction in oropharyngeal dose results from that large,

unrespirable aerosol particles are collected in the VHC rather than in the impactor throat

intake port when either VHC is attached. These data suggest that either VHC tested

would reduce the risk of topical adverse effects induced by fluticasone.

The highest respirable dose observed was from fluticasone propionate MDI alone.

However, the respirable dose from fluticasone propionate MDI alone was not

significantly larger than the respirable dose from the MDI attached to AeroChamberTM-

Plus or OptiChambero. The respirable doses from MDI alone, MDI attached to

AeroChamberTM-Plus or OptiChambers observed in this study were not significantly

different from previously reported values, respectively obtained in a nearly identical

manner [93, 94]. Our results confirmed that using AeroChamberTM-Plus or OptiChambero









in combination with a fluticasone propionate-1 10 MDI does not alter the respirable dose

available to the asthmatic patients. Further, the differences in delivered particle size

distribution of fluticasone aerosol, as reflected by both MMAD and GSD, between the

two VHCs and MDI alone were small. This suggests that fluticasone MDI can be used

successfully with AeroChamberTM-Plus or OptiChambero. However, the respirable dose

of fluticasone from OptiChambero'-Advantage was 40-45% less than from the MDI alone

or the AeroChamberTM-Plus VHC. Although the dose-response relationship in the

treatment of asthma has been difficult to establish, a 40-45% decrease in respirable dose

of inhaled fluticasone is likely to be a clinically relevant difference.

In vitro testing has some potential limitations. The most prominent of these is that

the correlation between the in vitro respirable dose of fluticasone and clinical response

has not been fully evaluated. Variability in patient inhalation technique and adherence

with the therapy are often the biggest determinants in clinical response and an in vitro

deposition study cannot account for these variables. Future work should include

measurement of aerosol deposition under more physiologic conditions likely to be

encountered in asthmatic patients. Another limitation of present study is that our results

are only reflective of the performance characteristics of the tested VHCs with the current

fluticasone MDI formulated with a chlorofluorocarbon propellant. Our results may not be

reflective of performance with other inhaled formulations of the same medication or of

different ICS. Although our study was conducted at only one flow rate (28.3 L/min),

other studies have shown that flow rates of 60 and 90 L/min have little effect on the

respirable dose from several corticosteroids MDIs when compared with that achieved at

30 L/min [21, 95].









There are differences in the in vitro performance characteristics of VHCs tested in

our study. Our results suggest that an asthmatic patient would receive the same dose of

fluticasone to the lung when using a MDI alone with proper inhalation technique or the

same MDI attached to AeroChamber-PlusTM Or OptiChambero. However, the same MDI

attached to OptiChambero'-Advantage, a newer VHC compared with the others in this

study, would result in 40-45% less fluticasone reaching the lung than that from the MDI

alone. This difference is likely to be clinically relevant. Hence, in asthmatics on

fluticasone MDI needing a VHC, AeroChamber-PlusTM Or OptiChambers would be a

better choice over OptiChambero'-Advantage since they do not change the respirable dose

compared to MDI alone.















CHAPTER 3
PLASMA CONCENTRATIONS AND PHARMACOKINETICS PROFILES OF
INHALED CORTICOSTEROIDS AND THEIR RELATION TO LUNG FUNCTION IN
ASTHMA

Introduction

As we discussed in a previous chapter, inhaled corticosteroids are absorbed into the

systemic circulation, predominantly from the lung and more variably from the

gastrointestinal tract. Therefore they have the potential to produce adverse systemic

effects. The risk of such adverse effects will depend on the extent of systemic (overall)

absorption of ICS and may differ between inhaled corticosteroids due to their different

physicochemical and pharmacokinetic properties. Recent studies have shown that

absorption and systemic effects of one inhaled corticosteroid, fluticasone propionate, are

greater in healthy volunteers than in subjects with asthma and airflow obstruction [54, 55,

96, 97]. However it was not true for budesonide [55, 98]. Another study showed that the

extent of adrenal suppression following 500Cpg fluticasone propionate was closely related

to lung function with greater cortisol suppression with increasing FEV1 [50]. Whether the

systemic absorption of other inhaled corticosteroids relates to lung function is unknown.

To explore this and to provide comparative data on the pharmacokinetic profiles of the

four drugs in the same subj ects, we compared plasma concentrations of beclometasone

monopropionate, budesonide, fluticasone propionate and mometasone furoate following

inhalation of a single dose in subj ects with asthma who had a wide range of FEV1 values.












Subj ects

Thirty non-smoking subj ects with asthma aged between 18 and 70 were recruited

between October 2003 and February 2004 from our volunteer database. Subj ects were

selected to provide a range of forced expiratory volume in one second (FEV1) values

from 30% predicted upwards. All had to have stable asthma defined as no change in

asthma symptoms or treatment for at least two months. There were no restrictions based

on treatment but subj ects were excluded if they had significant co-morbidity, were

pregnant or lactating or had a greater than 20 pack/year smoking history. The study was

approved by Nottingham Research Ethics Committee and written informed consent was

obtained from all subjects.

Measurements

Lung function

FEV1 and forced vital capacity (FVC) were measured with a dry bellows

spirometer (Vitalograph, Buckingham, UK) as the higher of two successive readings

within 100ml. The same spirometer was used for all measurements and was calibrated

weekly. Peak expiratory flow (PEF) was measured with a mini-Wright flowmeter as the

best of three readings.

Plasma drug assays

The plasma concentrations of budesonide, beclometasone monopropionate (the

active metabolite of beclometasone dipropionate), fluticasone propionate and

mometasone furoate were quantified with validated high performance liquid

chromatography/tandem mass spectrometry methods using a Micromass Quattro LC-Z

triple quadrupole mass spectrometer (Beverley, MA) [99-102]. The lower limits of


Methods









detection for the assays were 15 pg/ml for beclometasone monopropionate, fluticasone

propionate and mometasone furoate and 50 pg/ml for budesonide.

Protocol

Subj ects were screened at an initial visit to measure FEV1, assess their suitability

for the study and to familiarize themselves with the placebo dry powder inhalers.

Subj ects taking fluticasone propionate, budesonide or mometasone furoate were changed

to an equivalent dose of beclometasone dipropionate at least four days before the study

day; all subj ects were then asked to omit the beclometasone dipropionate from the

evening before the study day to allow time for beclometasone monopropionate to be

cleared from the plasma before the study. Subj ects attended the department in the

morning of the study day, a venous cannula was inserted and blood taken for baseline

drug assay. After ten minutes rest, FEV1, FVC and PEF were measured on two occasions

five minutes apart. Subj ects then inhaled two doses from each of four inhalers in random

order with a mouth rinse after each drug. The drugs and total doses administered were

beclometasone dipropionate 800Cpg (Becodisks@, Allen & Hanburys), budesonide 800Cpg

(Turbohaler@, Astra-Zeneca), fluticasone propionate 1000Cpg (Accuhaler@, Glaxo-

Wellcome) and mometasone furoate 800Cpg (Asmanex Twisthaler@, Schering-Plough).

Venous blood samples were taken into sodium fluoride/EDTA tubes at intervals over the

next eight hours, centrifuged at 1500 rpm for ten minutes and plasma samples extracted

and frozen at -70oC.

The main study endpoint was the relation between the subj ect' s% predicted FEVI

and the area under the plasma concentration/time curve (AUCo-s) for each drug. Power

calculations based on a linear regression model showed that 30 patients would give 90%

power to detect a correlation between FEV1 % predicted and AUC of 0.5 or more.










Analysis

Plasma concentrations of beclometasone monopropionate, budesonide, fluticasone

propionate and mometasone furoate were plotted against time for each subj ect;

pharmacokinetic parameters (maximum plasma concentration (Cmax), time to Cmax,

AUCo-s, mean residence time and terminal half-life) were calculated using a software

package (WinNonlin@ Professional Version 3.1 Pharsight Corporation, Mountain View,

CA). Concentrations below the limit of quantification were reported as missing and

extrapolated during non-compartmental analysis.

Plasma drug AUCo-s values were plotted against the mean of the two baseline FEV1

measurements, FEV1 % predicted and FEV1/FVC ratio, and against PEF and PEF %

predicted. The correlation coefficient (r) was calculated and trendlines fitted using linear

regression. Partial correlation coefficients were calculated for the relationship between

lung function and AUC for the four drugs after controlling for two potential confounding

factors, age and gender. To estimate the magnitude of the effect of lung function on AUC

values for each drug the equation derived from the linear regression model was used to

predict the AUC value at 100 and 50% predicted FEV1 and PEF and the ratio of these

two values was calculated.

Results

All 30 subjects (mean age 57, 16 men) completed the study as per protocol.

Baseline characteristics of the subj ects are shown in table 3-1. FEV1 prior to drug

inhalation ranged from 36 to 13 8% predicted. Four subj ects had detectable plasma

concentrations of one inhaled corticosteroid in their baseline sample. In two subj ects this

was beclometasone monopropionate (1152pg/ml and 46pg/ml), in one who usually took

budesonide it was budesonide (215pg/ml) and in one who usually took beclometasone

















Subj ect Sex Age

1 M 70
2 M 52
3 M 61
4 F 61
5 F 53
6 M 49
7 M 48
8 M 61
9 F 45
10 M 61
11 M 55
12 M 46
13 F 56
14 F 68
15 F 44
16 F 61
17 F 62
18 M 50
19 F 35
20 M 67
21 F 58
22 M 60
23 M 41
24 F 58
25 F 66
26 M 51
27 M 63
28 F 63
29 F 66
30 M 68
Mean 57
(SD) (9)


dipropionate, fluticasone propionate was detected (19pg/ml). Data for these drugs only

from the four subj ects were excluded from the analysis.


Table 3-1. Baseline characteristics


FEV1
(liters)
1.3
3.1
1.3
2.3
2.1
2.7
2.2
3.3
2.7
2.5
1.4
3.2
3.5
1.2
3.2
1.8
2.2
2.6
2.4
1.9
0.9
1.7
2.3
2.1
1.5
1.7
1.3
2.5
0.8
1.0
2.1
(0.8)


FEV1
(% pred)
48
90
36
98
86
78
66
106
100
82
42
74
138
62
113
72
96
75
93
65
47
53
57
87
77
46
43
99
44
40
74
(25)


PEF
%o pred)
50
79
42
97
79
75
79
96
90
78
58
82
101
62
95
70
107
70
97
71
45
40
71
83
72
67
48
110
43
43
73
(21)


FVC
(liters)
3.6
4.3
3.9
2.9
2.8
4.2
3.7
4
3.1
3.2
3
4.3
4.1
1.9
4.3
2.15
3
4.2
3.3
3.1
2
3.2
3.8
2.6
2.3
3.2
2.8
3.2
1.5
1.5
3.2
(0.8)


PEF
(liter/min)
270
480
245
425
360
460
360
555
425
450
350
530
460
255
455
310
465
430
460
390
195
235
455
370
300
415
275
485
180
180
376
(104)


(~


FEV1/FVC

0.35
0.71
0.32
0.79
0.74
0.65
0.59
0.82
0.88
0.78
0.47
0.74
0.86
0.63
0.74
0.81
0.72
0.63
0.73
0.61
0.44
0.53
0.59
0.78
0.66
0.54
0.46
0.8
0.55
0.55
0.6
(0.2)










Plasma drug concentrations and pharmacokinetic profiles

All four inhaled corticosteroids were present at detectable plasma concentrations

following inhalation in all subj ects and plasma concentrations were above the lower limit

of quantification in the maj ority of samples (99%, 97%, 95% and 93% of samples for

budesonide, beclometasone monopropionate, fluticasone propionate and mometasone

furoate, respectively).




0.4. 2
BMP BUD


0.2. 1









ca0.05. 0.05



0.:: 0.:
0 2 4 6 8 0 2 4 6 8

Time (hours)



Figure 3-1. Mean (SE) plasma concentration-time curves for beclometasone
monopropionate (BMP), budesonide (BUD), fluticasone propionate (FP) and
mometasone furoate (MF) following inhalation (Note that Y axes have
different scales)

The shape of the mean plasma drug concentration-time curves and pharmacokinetic

parameters varied considerably between drugs as shown in figure 3-1 and table 3-2. Data

from two subj ects for one inhaled corticosteroid (one beclometasone dipropionate, one









mometasone furoate) were unsuitable for pharmacokinetic modelling and excluded from

the calculations.


Table 3-2. Mean (SD) pharmacokinetic parameters for beclometasone
monopropionate(BMP),budesonide(BUD), fluticasone propionate(FP) and
mometasone furoate (MF)
BMP BUD FP MF
Tmax (h) 2.93 (0.87) 0.16 (0.12) 1.19 (0.91) 1.28 (0.96)
MRT (h) 12.3 (13.9) 3.6 (0.8) 11.4 (14.6) 8.4 (5.8)
T1/2 7.6 (9.7) 2.7 (0.5) 7.8 (10.0) 5.6 (4.1)
Cmax(ng/ml) 0.34 (0.15) 1.59 (0.88) 0.10 (0.04) 0.09 (0.06)
AUC(ng/ml/h) 1.79 (0.78) 3.43 (1.29) 0.44 (0.23) 0.43 (0.32)

Peak plasma concentrations following inhalation occurred within 10 minutes with

budesonide, and after 1.2, 1.3 and 2.9 hours with fluticasone propionate, mometasone

furoate and beclometasone monopropionate respectively. Mean pulmonary residence

times showed a similarly wide variation from 3.6 hours with budesonide to 12.3 hours

with beclometasone monopropionate. The terminal half-life ranged from 2.7 hours for

budesonide to 7.8 hours for fluticasone. The magnitude of the peak plasma concentration

also differed markedly between drugs being highest for budesonide at 1.6 ng.ml-l and 18,

16 and 5 times lower for mometasone furoate, fluticasone propionate and belcometasone

monopropionate, respectively. AUC values ranged from 3.43 ng/ml/h for budesonide to

0.43 ng/ml/h for mometasone furoate (table 3-2).

Relation between lung function and AUC

AUCo-s values tended to be higher in patients with a higher FEV1 % predicted for

all four drugs as can be seen in figure 3-2. The correlation coefficients (r) for the relation

between FEV1 % predicted and AUC values for fluticasone propionate, budesonide,

beclometasone monopropionate, and mometasone furoate were 0.30 (p = 0. 11), 0.36 (p =

0.05), 0.39 (p = 0.04) and 0.55 (p = 0.002) respectively. The same pattern was seen for






































.


MF


* .


the relation between PEF % predicted and AUC values with r values for fluticasone

propionate, budesonide, beclometasone monopropionate, and mometasone furoate being

0.26 (p = 0.17), 0.35 (p = 0.07), 0.42 (p = 0.03) and 0.60 (p < 0.001) respectively. The

correlation between other measures of lung function (absolute FEV1, PEF and FEV1/FVC

ratio) and AUC was similar but generally weaker than that between FEV1 and PEF %

predicted and AUC. Both age and gender caused a 10% or greater change in the

correlation coefficients. The correlation between FEV1% predicted and AUC was

stronger after adjusting for age but weaker after adjusting for gender (Table 3-3).


100 150 0 50

FEVi % predicted


100 150


Figure 3-2. Relation between FEV1% predicted and the area under the plasma
concentration-time curve (AUC) for beclometasone monopropionate (BMP),
budesonide (BUD), fluticasone propionate (FP) and mometasone furoate (MF)
following inhalation (Note that Y axes have different scales)


BMP


BUD











Table 3-3. Partial correlation coefficients (r) for the relationship between FEV1 %
predicated and AUC after adjusting for age or gender
r p value
Beclometasone monopropionate Unadjusted 0.39 0.04
Age 0.48 0.01
Gender 0.35 0.08
Budesonide Unadjusted 0.36 0.05
Age 0.37 0.05
Gender 0.28 0.15
Fluticasone propionate Unadjusted 0.3 0.11
Age 0.45 0.01
Gender 0.17 0.39
Mometasone furoate Unadjusted 0.55 0.002
Age 0.63 < 0.001
Gender 0.48 0.009

Estimated AUC values were higher at 100 % compared to 50% predicted FEVI

for all four drugs, the ratios for budesonide, fluticasone propionate, beclometasone

monopropionate and mometasone furoate being 1.3, 1.4, 1.4 and 2.3 respectively. The

same pattern was seen with PEF % predicted (Table 3-4).


Table 3-4. Predicted AUC values at 50 and 100% predicted FEV1 and PEF and the ratio
of these values
AUC at 50% AUC at 100% Ratio of AUC
predicted predicted values
Beclometasone FEV1 1.52 2.1 1.4
monopropionate PEF 1.45 2.22 1.5
Budesonide FEV1 3.01 3.93 1.3
PEF 2.95 4.08 1.4
Fluticasone FEV1 0.37 0.51 1.4
propionate PEF 0.37 0.51 1.4
Mometasone FEV1 0.27 0.61 2.3
furoate PEF 0.23 0.68 3


Discussion

This is the first study to compare plasma concentration-time curves and

pharmacokinetic profiles ofbeclometasone monopropionate, budesonide, fluticasone









propionate and mometasone furoate following inhalation in the same patients and to

explore the extent to which systemic absorption of all four drugs relates to lung function.

Plasma Drug concentrations and Pharmacokinetic Profiles

The plasma drug concentration-time curves and pharmacokinetic profiles of

fluticasone propionate and mometasone furoate were broadly similar to each other but

differed markedly from the other two drugs, beclometasone monopropionate and

budesonide. Budesonide, the most water-soluble of the four drugs, had the shortest

pulmonary residence time, time to peak plasma concentration (Tmax) and terminal half-

life; beclometasone monopropionate had the longest pulmonary residence time and Tmax

and its terminal half-life was also longer and similar to that of fluticasone propionate.

The pharmacokinetic parameters determined for budesonide are in agreement with

previous data [55, 96, 97, 103, 104] and the rapid absorption is related to its lower

lipophilicity and greater solubility in bronchial fluid [105]. The shorter terminal half-life

of budesonide primarily reflects its lower volume of distribution since high clearance is a

feature of all the inhaled corticosteroid[106] Budesonide had a much higher peak plasma

concentration than the three other drugs, an effect attributed to high pulmonary

deposition by the Turbohaler@, its rapid absorption and low volume of distribution [104].

The long pulmonary residence time and Tmax with beclometasone monopropionate

reflects its greater lipophilicity and greater oral bioavailability (41% compared to 11% for

budesonide and <1% for fluticasone propionate and mometasone furoate [80, 106-108])

since absorption from the gastrointestinal tract is slower than that from the lungs [107].

Our values for pulmonary residence time and terminal half-life for beclometasone

monopropionate were longer than those reported in studies using metered dose inhalers,

[109-111] perhaps due to differences in particle characteristics and deposition pattern









with dry powder inhalers. The lower peak plasma concentrations for beclometasone

monopropionate compared to budesonide is related to its slower absorption and greater

volume of di stribution.

Fluticasone propionate and mometasone furoate had fairly similar absorption

profiles. The long terminal half-life of fluticasone propionate may be due to its large

volume of distribution or 'flip-flop pharmacokinetics' due to slow release from the lungs.

Both drugs had particularly low peak plasma concentrations, even though fluticasone

propionate was given at a slightly higher dose than the other three inhaled corticosteroids.

Slow absorption from the lungs, increased mucociliary clearance and high volumes of

distribution are likely to explain these findings. Our pharmacokinetic data for fluticasone

propionate are similar to those reported previously [55, 96, 97, 112, 113], whilst the only

published data on mometasone furoate are difficult to compare since a relatively

insensitive assay was used [114].

Relation Between Lung Function and AUC

We found that AUC values tended to be higher in patients with better lung function

for all four drugs. Our estimates suggest that the systemic absorption of budesonide,

fluticasone propionate, beclometasone monopropionate and mometasone furoate is some

1.3, 1.4, 1.4 and 2.3 times higher at 100% predicted FEV1 compared to 50% predicted

FEV1. The effect appears to be greatest for mometasone furoate although the study was

not powered to compare the magnitude of effect between drugs.

The inhaled corticosteroids were delivered by dry powder inhalers in our study, and

reduced inspiratory flow may have caused less drug to be delivered to the airways in

subj ects with greater airflow obstruction. Furthermore, the drug that enters the airways is

more likely to be deposited in central airways in patients with airflow obstruction, and









hence more likely to be removed by mucociliary clearance. This would be expected to

affect drugs that dissolve slowly in airway lining fluid such as mometasone furoate and

fluticasone propionate most and drugs with high water solubility, such as budesonide,

least [105].

When we explored the effect of age and gender on the relation between lung

function and AUC values we found the relation to be stronger after adjusting for age but

weaker after adjusting for gender. This highlights the fact that factors other than lung

function affect the systemic absorption of inhaled corticosteroids thus reducing the

strength of the relation between lung function and plasma concentration AUC values in

cross-sectional studies. This conclusion is supported by our later study, presented in next

chapter, in which we compared plasma concentrations of budesonide and fluticasone

propionate following inhalation with and without prior methacholine-induced

bronchoconstriction (mean fall in FEV1 of 33%). Mean AUC values were reduced for

both drugs after lowering FEV1, i.e. when we looked at within subject changes, but not

when we compared differences in FEV1 on AUC between subj ects. Studies of the effect

of lung function on the systemic absorption of inhaled corticosteroids across populations

need to allow for between-subj ect variability.

Conclusion

In conclusion our study has shown marked differences in plasma concentration-

time curves and pharmacokinetic profiles between beclometasone monopropionate,

budesonide, fluticasone propionate and mometasone furoate following inhalation. The net

effect of these differences on the benefit to risk ratio of inhaled corticosteroids is complex

and needs to be explored further. The systemic absorption of all four inhaled

corticosteroids appears to be greater in patients with a higher FEV1 % predicted although






44


other factors, including age and gender affect this relationship. Our findings re-enforce

the importance of reviewing the need for higher doses of an inhaled corticosteroid,

particularly in patients with relatively normal lung function and in patients whose lung

function improves with treatment.















CHAPTER 4
FLUTICASONE AND BUDESONIDE CONCENTRATIONS AFTER INTHALATION
EFFECT OF BRONCHOCONSTRICTION

Introduction

Adverse systemic effects from inhaled corticosteroids have been recognized

increasingly over recent years. With around 5 % of the population in more economically

developed countries prescribed an inhaled corticosteroid, often for many decades,

understanding the factors that determine these adverse effects is important. All currently

available inhaled corticosteroids are absorbed into the systemic circulation and hence

have the potential to cause adverse systemic effects. The risk may differ between inhaled

corticosteroids, however, due to their different physicochemical and pharmacokinetic

properties and there is evidence that these drug related differences interact with patient

factors such as airflow obstruction [115]. Plasma drug concentrations following

inhalation of 1000 Cpg fluticasone were considerably lower in people with airflow

obstruction than in healthy volunteers but this was not the case for budesonide [51, 55,

97]. The differences between fluticasone and budesonide were attributed to differences in

lipophilicity between the two drugs [55]. Whether changes in airflow obstruction within

an individual affect systemic absorption in the same way is unknown, but important in

view of the fluctuations in airway caliber seen in asthma. To explore this we compared

the plasma concentrations of fluticasone and budesonide following inhalation of a single

dose, with and without prior methacholine-induced bronchoconstriction.









Methods

Subj ects

Twenty non-smoking subj ects with asthma aged between 18 and 70 were recruited

from our volunteer database. To be included subjects had to have a forced expiratory

volume in one second (FEV1) of at least 80 % predicted and 1.5 litres, a provocative dose

of methacholine causing a 20 % fall in FEV1 (PD20) below 8 CLM, and stable asthma,

defined as no change in asthma symptoms or treatment for two months. Subj ects were

excluded if they had significant co-morbidity, were taking any medication known to alter

the metabolism of corticosteroids, were pregnant or lactating, or had a greater than 20

pack year smoking history. The study was approved by Nottingham Research Ethics

Committee and written informed consent was obtained from all subjects.

Measurements

Spirometry

FEV1 and forced vital capacity (FVC) were measured with a dry bellows

spirometer (Vitalograph, Buckingham, UK) as the higher of two successive readings

within 100 ml.

Methacholine inhalation challenge

Subj ects inhaled three puffs of normal saline followed by doubling doses of

methacholine from 0.048 CLM to a maximum of 24.5 CLM from DeVilbiss nebulizers and

FEV1 was measured one minute after each dose [116]. The test was stopped once FEV1

had fallen by 20 % from the post-saline value during screening, for calculation of PD20

by interpolation, and once FEV1 had fallen by 25 % in the main study.










Drug assays

Plasma concentrations of fluticasone propionate and budesonide were quantified

with previously validated high performance liquid chromatography/tandem mass

spectrometry methods using a Micromass Quattro LC-A triple quadrupole mass

spectrometer (Beverley, MA) [99, 100]. The lower limits of detection for the assays were

15 pg/ml for fluticasone propionate and 50 pg/ml for budesonide. The intra- and inter-

assay coefficients of variation for the assays were below 13.6 for both fluticasone and

budesonide at concentrations between 0.015 and 0.75 ng/ml for fluticasone and between

0.15 and 2.5 ng/ml for budesonide.

Protocol

Subj ects were screened to assess suitability and to measure FEV1 and PD20

methacholine. Subj ects familiarized themselves with the two dry powder inhalers

(Accuhaler@, Glaxo-Wellcome and Turbohaler@, Astra-Zeneca) to ensure optimal

inhaler technique during the study and those taking fluticasone or budesonide were

changed to an equivalent dose of beclometasone dipropionate for four days before and

until study completion. Subjects were asked to avoid short and long acting p2 agOnists for

12 hours, and to avoid exercise and caffeine on the morning of the study.

Subj ects attended for two study visits at the same time of day & one hour, 7 & 3

days apart. A venous cannula was inserted and after ten minutes rest, FEV1 was measured.

1000 Cpg fluticasone (Accuhaler@), given as two 500 Cpg doses, and 800 Cpg budesonide

(Turbohaler@), given as two 400 Cpg doses, were then inhaled in random order; after each

drug subj ects rinsed their mouths with water which was then discarded. Venous blood

samples were taken into heparinised tubes at intervals over the next 5 hours (0, 5, 15, 30,

60, and 90 minutes, 2, 3, 4 and 5 hours), centrifuged at 1500 rpm for 10 minutes and










plasma samples were then frozen at -70 oC. The protocol for the two study visits was

identical except that on one occasion prior to inhalation of the drugs, a methacholine

challenge was carried out to induce a fall in FEV1 of at least 25 % from baseline FEV1.

The two studies were carried out in random order according to a computer generated code.

Power calculations showed that 20 patients would give 90 % power to detect a 0.7

SD difference in the area under the plasma concentration time curve over five hours, the

primary outcome, between the two study visits.

Analysis

Plasma fluticasone and budesonide concentrations were plotted against time for

each subj ect and the maximum plasma concentration (Cmax), time to maximum plasma

concentration (Tmax) and area under the curve up to five hours (AUCo-S) calculated using

a software program (WinNonlin@ Professional Version 3.1, Pharsight Corporation,

Mountain View, CA). Cmax, Tmax and AUCo-S values for the two study visits were

compared for each drug by paired t-test. To allow for the differences in concentrations of

the two drugs we calculated the ratio of the AUCo-s values on the days with and without

bronchoconstriction for each subj ect for each drug and compared the values between

drugs by Wilcoxon signed rank test. This analysis was repeated for Cmax values.

Results

All 20 subj ects (mean age 48, 12 men) completed the study and their baseline

characteristics are shown in table 4-1. All were non-smokers and only two had ever

smoked.

Subj ects had a geometric mean PD20 methacholine of 1.2 pLM (range 0.06 to 6.1i).

Mean (SD) FEV1 was 2.9 (0.7) litres (91 % predicted) on screening and 2.8 (0.7) and 2.9

(0.6) litres on arrival for the visits with and without bronchoconstriction. All subj ects had










a fall in FEV1 of at least 25 % (range 25 to 47 %) following methacholine giving an FEVI

of 1.9 (0.5) litres (60 % predicted) prior to drug inhalation and a mean difference in FEV1

prior to drug inhalation on the two days of 0.9 litres (range 0.5 to 1.35) or 33 % (range 20

to 45 %).


Table 4-1. Baseline characteristics


* Beclometasone dipropionate
** Geometric mean


Data for budesonide plasma concentrations were lost for two subj ects due to a

computer problem during drug analysis. Complete data were therefore available for 20


Subject

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Mean
(SD)


Sex Age Height
(cm)
M 55 192
M 50 173
F 50 160
M 63 174
F 67 155
M 23 170
F 39 161
M 38 189
M 41 182
M 41 176
M 52 172
F 62 166
M 66 173
M 48 166
M 49 164
M 42 179
F 65 161
F 27 157
F 34 161
F 44 166
48 167
(13) (10)


Weight
(kg)
78
78
86
77
68
79
63
92
96
87
88
105
78
79
68
67
64
57
71
60
77
(13)


FEV1
(liters)
3.8
2.8
2.55
2.8
1.55
4
3.5
3.8
3.8
3.8
3.05
1.95
2.4
2.75
2.8
3.55
2
2.8
2.9
2.35
2.9
(0.7)


FEV1
% pred
91
80
103
89
84
96
126
84
92
98
90
81
80
84
89
89
93
96
100
82
91
(11)


PD20
(micromol)
2.7
3.4
5.7
2.7
0.1
2.3
0.5
0.9
0.4
5
0.1
0.06
6.1
1
5.1
0.4
6.1
2.2
1.1
0.6
1.2**


Usual ICS

BDP*
None
BDP
BDP
None
None
BDP
None
BUD
None
None
BDP
BDP
FP
BUD
BDP
None
FP
FP
FP


subj ects for fluticasone and 18 subj ects for budesonide.










The shape of the fluticasone and budesonide plasma concentration-time curves

showed marked differences on the study day without bronchoconstriction. Peak plasma

concentrations of budesonide were 14 times higher than those of fluticasone and occurred

considerably earlier as shown in figure 4-1; mean (SD) time to maximum concentration

(Tmax) values were 0.21 (0.24) hr and 1.21 (0.91) hr for budesonide and fluticasone

respectively. The shape and time course of the plasma drug concentration curves for each

drug were similar on the two study days although plasma levels were lower following

methacholine-induced bronchoconstriction (figure 4-1 and table 4-2).


Table 4-2. Pharmacokinetic parameters for fluticasone and budesonide when inhaled with
and without prior methacholine-induced bronchoconstriction
Fluticasone Budesonide
Without Following Without Following
bronchoconstriction bronchoconstriction bronchoconstriction bronchoconstriction
(SD) (SD) (SD) (SD)
Cmax (ng.ml )> 0.12 (0.07) 0.06 (0.02) 1.67 (0.82) 1.16 (0.48)
Tmax (h) 1.21 (0.91) 0.78 (0.62) 0.21 (0.24) 0.23 (0.24)
MRT (h) 7.79 (8.13) 4.88 (1.79) 3.09 (1.09) 3.17 (1.73)
Cmax Maximum plasma concentration
Tmax Time of maximum plasma concentration
MRT Medium residence time

Mean Cmax values for both fluticasone and budesonide were lower when the drugs

were inhaled following bronchoconstriction (table 4-2). Cmax values were a median 44 %

lower (interquartile range (IQR) 21 to 60) for fluticasone and 33 % lower (IQR -7 to 51)

for budesonide following bronchoconstriction but the ratio between study visits for

fluticasone was not significantly different to the ratio for budesonide, p=0.18.

Mean AUCo-s values, the primary endpoints, for both fluticasone and budesonide

were lower when the drugs were inhaled following bronchoconstriction (table 4-3).

AUCo-5 values were a median 60 % lower (IQR 36 to 75) for fluticasone and 29 % lower

(IQR 2 to 44) for budesonide following bronchoconstriction. The ratio for AUCo-s values






51


between study visits for fluticasone was significantly greater than the ratio for budesonide,

p=0.007.


O 1 2 3 4 5


Figure 4-1. Mean (SE) plasma drug concentrations following inhalation of 1000 Cpg
fluticasone (Accuhaler@) and 800 Cpg budesonide (Turbohaler@) in 20
subj ects with asthma with ( +-) and without (-a) prior methacholine-
induced bronchoconstriction (Note: Y axes have different scales).


Discussion

This is the first study to explore the effect of change in airflow obstruction within

an individual on plasma concentrations of fluticasone and budesonide following drug










inhalation. There was an average 33 % difference in FEV1 prior to drug inhalation on the

two study visits as a result of methacholine challenge. Plasma concentrations of both

fluticasone and budesonide were lower when FEV1 was reduced but the magnitude of this

effect was greater for fluticasone.


Table 4-3. Area under the curve (AUC) values for fluticasone and budesonide when
inhaled with and without prior methacholine-induced bronchoconstriction
Fluticasone Budesonide
Without Following Differences Without Following Differences
broncho- broncho- (95% CI), p broncho- broncho- (95% CI), p
constriction constriction value constriction constriction value
(SD) (SD) (SD) (SD)
AUC 0.40 0.16 0.23 2.87 2.28 0.60
(ng.ml '.h) (0.23) (0.10) (0.14 to 0.33), (1.19) (0.90) (0.02 to 1.17),
(SD) p <0.001 p =0.04

The greater effect of bronchoconstriction on plasma concentrations of fluticasone

compared to budesonide is similar to the findings in patients with and without airflow

obstruction [55, 96, 97, 105]. Bronchoconstriction will cause a greater proportion of both

drugs to be deposited in more central airways, where mucociliary clearance is better able

to remove drug [117]. This would be expected to affect fluticasone more than budesonide

because fluticasone, being more lipophilic, dissolves more slowly in airway lining fluid

and hence is available for clearance for a longer time [105]. Furthermore little of the

fluticasone that is removed by mucociliary clearance will reach the systemic circulation

due to its low oral bioavailability (<1 % compared to around 10 % for budesonide) [105].

Peak plasma drug concentrations occurred earlier and were considerably higher

following inhalation of budesonide than fluticasone in keeping with previous data [55].

These differences are likely to reflect high pulmonary deposition of budesonide from the

Turbohaler@, its higher oral bioavailability, and greater water solubility and its lower

volume of distribution [105, 115, 118].










We gave fluticasone and budesonide at the same time to ensure that they were

studied under identical conditions. This approach was validated by Agertoft and Pederson

who found similar plasma fluticasone and budesonide concentrations following

inhalation whether the drugs were given separately or together [119]. Drug

concentrations were measured over five hours since our previous study showed that the

maj or differences in plasma concentrations between subj ects with and without airflow

obstruction occurred during this time [55]. We did not seek to obtain pharmacokinetic

parameters other than Cmax and Tmax, since these are already available and a longer

sampling period would have been required [115]. We believe our findings are due to

reduced rather than delayed, drug absorption in view of the shape of the plasma

concentration-time curves and the fact that the changes within subj ects in this study are

very similar to the differences seen over 8 and 12 hours between subj ects with and

without airflow obstruction[51, 55, 97].

By using a methacholine challenge we were able to study the effect of change in

lung function within an individual on plasma concentrations of fluticasone and

budesonide under controlled conditions. This is likely to be a reasonable model for

changes in airflow that occur as asthma control deteriorates. A 33 % reduction in FEV1 in

our study caused a 60 % reduction in AUCo-s values for fluticasone suggesting a 60 %

reduction in systemic exposure. Equally, as FEV1 increases with treatment, systemic

exposure would be expected to increase in a similar way. The risk of adverse systemic

effects from inhaled fluticasone is likely, therefore, to vary considerably in relation to

lung function within an individual while that from budesonide is unlikely to be affected

to the same extent. These findings re-enforce the importance of reviewing the need for






54


higher doses of inhaled corticosteroids and particularly fluticasone, as lung function

improves and in patients with relatively normal lung function.















CHAPTER 5
EVALUATION OF THE ADMINISTRATION TIMVE EFFECT ON THE
CUMULATIVE CORTISOL SUPPRESSION AND CUMULATIVE LYMPHOCYTES
SUPPRESSION FOR ONCE-DAILY INHALED CORTICOSTEROIDS

Introduction

Inhaled corticosteroids (ICS) continue to remain the first-line therapy for the

treatment of asthma [12]. Clinical studies have demonstrated the high anti-asthmatic

efficacy (e.g. prevention and control of symptoms) and reduced systemic side effects of

ICS therapy. However, potential systemic side effects induced by orally (swallowed) or

pulmonary absorbed drug remain a concern in the eyes of the patient [13, 64, 69]. The

maj or systemic effects include bone demineralization, growth suppression in children,

hypothalamic-pituitary-adrenal (HPA) axis suppression and etc sec. Since many of these

effects are either difficult to quantify or not possible to measure during short-term

observations, the use of biomarkers is common in clinical pharmacology studies of

corticosteroids. Two such markers have been identified for the systemic effects of

corticosteroids: suppression of endogenous cortisol and blood lymphocytes [120, 121].

Cortisol serum levels are very sensitive to the presence of exogenous

corticosteroids and the release of serum cortisol is often dramatically suppressed after

drug administration. Thus, cortisol suppression as monitored by serum or urinary cortisol

is currently used in clinical pharmacology studies for the assessment of systemic side

effects, as these parameters are sensitive and relatively easy to determine while other

more clinically relevant parameters (e.g. growth suppression) is used less frequently in

specific patient populations. Twenty-four hour cortisol suppression has been used as the









most sensitive marker [122] and adequate PK/PD models have been developed [123].

Endogenous serum cortisol concentrations follow a circadian rhythm. Peak concentration

is reached in the morning between 6AM and 10 AM, trough concentrations are reached at

night between 8PM and 2AM [124-126]. Therefore, cortisol suppression after ICS is

expected to undergo diurnal variation dependent on the administration time in the day

[127].

It is known that the administration of corticosteroids results in a transient depletion

of blood lymphocytes, which is called lymphocytopenia [128, 129]. Lymphocytopenia

seems to be an appropriate choice as biomarker since the decline in lymphocytes is

directly related to the suppression of immune system often observed in chronic

corticosteroid therapy, whether it is side effect or desired outcome [130-134].

Endogenous cortisol, like exogenous corticosteroids, also affects the number of

lymphocytes in the blood. Since the endogenous serum cortisol concentration follows a

circadian rhythm, in the absence of additional stimuli the lymphocytes count shows

intraday fluctuations inversely related to cortisol concentrations [121]. The overall extent

of lymphocytopenia observed after exogenous corticosteroid administration is the

combined effect of a direct suppression of blood lymphocytes by the exogenous

glucocorticoid and a reduction of this effect due to the decrease in serum cortisol level

induced by exogenous glucocorticoids. Therefore, the extent of lymphocytes suppression

in the blood after ICS administration may also undergo diurnal variation dependent on

the administration time of the ICS during the day. The importance of the administration

time of corticosteroids for lymphocytes suppression has not yet been recognized.









Newer inhaled glucocorticoids are currently being dosed once a day. Clinical

studies in recent years have tried to evaluate the effect of morning or evening dosing on

the effect and systemic side effect with contradicting results [41-48] In this study, a

population modeling/simulation approach was developed and applied to evaluate the

effect of daily administration time on 24-hour cumulative cortisol suppression (CCS) and

cumulative lymphocytes suppression (CLS).

Methods

Historic Data

Results of a previously published study comparing the single-dose and steady-state

(day 5) pharmacokinetics and pharmacodynamics of inhaled fluticasone propionate and

budesonide in healthy subj ects were used to develop the necessary population

pharmacokinetic and pharmacodynamic model [103]. The study was conducted in

accordance with the revised Declaration of Helsinki (Hongkong Revision 1989) and to

Good Clinical Practice guidelines. A total of 14 healthy male volunteers, at mean age of

26.4 years (range 22-32), height of 179.2 (range 172-187) cm and weight of 72.7 (range

62-85) kg completed the study. The study evaluated the single dose and steady state

pharmacokinetic and pharmacodynamic properties for inhaled fluticasone propionate

(200 and 500 Clg via Diskuse, FP-DSK) and budesonide (400 and 1000 Clg via

Turbohalero, BUD-TBH) in healthy subj ects. During this double blind, double dummy,

randomized, placebo controlled, five-way crossover study single doses were administered

at 8 am on day 1 followed by twice-daily dosing at 8 a.m. and 8 p.m. on days 2-5. During

day 1 and 5 blood samples for drug, cortisol and lymphocytes analysis were collected at









frequent intervals. On days 2-4, only trough samples (8 a.m. and 8 p.m.) were obtained.

The detail of the study has been reported elsewhere [103].

Population Pharmacokinetic/Pharmacodynamic Analysis

A sequential modeling approach was taken to facilitate the data analysis. In the first

step, the population pharmacokinetic analysis was performed. This was followed by the

population pharmacodynamic analysis with individual pharmacokinetic parameter

estimates as part of input. Log-transformed data were used in the analysis since it was

found that log-transformation of the data stabilized the model. The analysis was carried

out using the NONMEM program (Version V, Globomax, Hanover, Md).

Pharmacokinetic analysis
Based on previously performed standard two-stage pharmacokinetic analysis, 1 and

2 compartment body model with first-order absorption were both tried to identify the

optima structure model for BUD and FP, respectively [135, 136]. The inter-subject

variability of PK parameters was modeled as exponential error model, as follows:

P, = 0 -exp(r,)

where Pi is the parameter for ith subj ect, 6 is the typical population value of the parameter,

and r1i is a random inter-subj ect effect with mean 0 and variance Co2. The intra-subj ect

variability was modeled as additive error model, as follows:

In(y,)= In~yp)+ S

where yij and ypij represent the ith subj ect' s jth observed and predicted concentration,

respectively, and a is the random residual error, which is normally distributed with mean

0 and variance G2. Fist-order conditional method (FOCE) was used for parameter

estimation. Individual subj ect PK parameters were calculated using the posterior

conditional estimation technique of NONMEM. After determination of structure










model, inter-occasion variability was explored for all PK parameters [137]. The

significance level required to keep inter-occasion variability on a parameter in a

hierarchical model was set to p < 0.01, which corresponds to a drop in the obj ective

function by 6.64. Finally, a bootstrap resampling technique was applied as internal

validation of the model [138]. Wings for NONMEM (Version 3.04, developed by Dr. N.

Holford) was used for execution of bootstrap resampling analysis.

Pharmacokinetic/Pharmacodynamic analysis of effects of BUD and FP on cortisol

In the absence of exogenous stimuli, the circadian rhythm of endogenous cortisol

was describe by a linear release model, as follows [139]:


Rc = mx-t Ra- tmn tmax < t < tmin
Vd (t,, t,, 24) Vd (t,, t,, 24)


Rc = max t_ max min ,tmin < t < tmax
Vd -(t -t ) d(t t)
max mmmax m

dCIcort
0 = Rc -kco'' CcttOr,


where Rc is the release rate of cortisol, Rmax is the maximum cortisol release rate, is the

time of maximum release, is the time of minimum release of cortisol, Vd is the volume of

distribution of cortisol and fixed at literature value 33.7 L, Cttco"~is the cortisol

concentration, Ke is the elimination rate constant of cortisol and fixed at literature value

0.56 h- The change in total cortisol plasma concentration and its modulation by the free

plasma concentration of an exogenous corticosteroid was described with an indirect

response model, as follows [140]:










dC, "O' Em C'ECS -Cort Cor
= R -1 k -
dt c EC-ECS4Cort C-ECS e tot


where Emax is the maximum suppressive effect and fixed at 1, C ECS is the unbound

exogenous corticosteroid concentration and E50ECS'Codt is the unbound exogenous

corticosteroid concentration to achieve 50% of the maximum suppression [139, 140]. The

unbound faction in plasma is 0.12 for BUD, and 0.1 for FP [80].

The inter-subj ect variability of PD parameters was modeled as exponential error

model, as described in previous section, and residual error was modeled as additive error

model, as described in previous section. Fist-order (FO) method was used for parameter

estimation. Individual subj ect PD parameters were obtained as empirical Bayes estimates

using the POSTHOC option of NONMEM. The cortisol data after placebo were used to

calculate tmax and tmin. The cortisol data after active treatment were modeled with PK

parameters, tmax and tmin fixed to estimate Rmax and EC50ECS'Cort

Inter-occasion variability was explored for all PD parameters. The significance

level required to keep inter-occasion variability on a parameter in a hierarchical model

was set to p < 0.01, which corresponds to a drop in the objective function by 6.64. The

final model was validated by prediction check: the serum cortisol concentrations were

simulated 100 times, then 24-hour CCS at steady state and observed 24-hour CCS at

steady state were compared using Kruskal-Wallis test followed by Dunn's multiple

comparison. The 24-hour CCS is quantified as follows [127]:

A LE rSupp, 24 A LEBase, 24 A LETherapv, 24
%CC'S = ~
A UC Base,24 A UC Base,24










where AUCSupp,24 iS the difference between the area under the cortisol plasma

concentration-time curve at baseline (AUCbase,24) and during therapy (AUCtherapy,24) OVeT

24 hour.

Pharmacokinetic/Pharmacodynamic analysis of effects of budesonide and
fluticasone propionate on lymphocytes

The combined pharmacodynamic effects of cortisol and exogenous corticosteroid

on lymphocyte can be described as [141]:


E CECS E50 Corty
dN max EC50CortoLym /
=~~ K 1 -k N
dt mE C y u
ECECS4Lym + ECS E50 Corty
50 f EC50CortoLym /


where N is the number of lymphocytes in blood, Cc"o" is the unbound cortisol

concentration, Kin is the influx rate constant of cells from the extravascular space, Kout is

the efflux rate constant of cells out of the vascular space, Emax is the maximum effect

and fixed at 1, E50CortaLym is the unbound concentration of cortisol to produce 50% of the

maximum effect and E50ECS'Lym is the unbound concentration of to achieve 50% of the

maximum effect. Cortisol exhibits non-linear binding to serum protein in the

physiological concentration range and the unbound concentration can be calculated by

solving following equation [120]:


Cor / +K~lb X Ab! X C~ort+ Cort
tr 1+ Kro x CICOrt f


where KTc is the affinity constant between cortisol and transcortin (3x107 1/mol), KAlb the

affinity constant between cortisol and albumin (5000 1/mol), QTc the total transcortin

concentration (0.7 Clmol/1), QAlb the total albumin concentration (550 Clmol/1) and Ccort

the unbound cortisol serum concentration [142].









The inter-subj ect variability of PD parameters was modeled as exponential error

model, and residual error was modeled as additive error model. Fist-order (FO) method

was used for parameter estimation. Individual subj ect PD parameters were obtained as

empirical Bayes estimates using the POSTHOC option of NONMEM. The lymphocyte

data were modeled to estimate Kin, Kout, EC50CortaLyn' and EC50ECS4Lyn' with all the other

PK and PD parameters fixed.

Inter-occasion variability was explored for all PD parameters. The significance

level required to keep inter-occasion variability on a parameter in a hierarchical model

was set to p < 0.01, which corresponds to a drop in the objective function by 6.64. The

final model was validated by prediction check: the blood lymphocyte counts were

simulated 100 times, then 24-hour CLS at steady state and observed 24-hour CLS at

steady state were compared using Kruskal-Wallis test followed by Dunn's multiple

comparison. The 24-hour CLS is quantified as follows [127]:


%C'LS = ~


where AUCSupp,24 iS the difference between the area under the lymphocytes plasma

number of counts time curve at baseline (AUCbase,24) and during therapy (AUCtherapy,24)

over 24hr.

Population PK/PD Simulation to Evaluate Administration Time Effect on CCS and
CLS for Once-daily Dose of BUD and FP

Using the developed population PK/PD model, we performed simulations to

evaluate the effect of the administration time effect on 24-hour CCS and 24-hour CLS at

steady state for 4 clinically relevant dosing regimens: Img once-daily BUD, 2mg once-

daily BUD, 0.5mg once-daily FP and Img once-daily FP. For each dosing regimen,









simulation was performed at 24-hourly distributed administration time points throughout

the day. NONEME was used for execution of the simulation.75 hypothetic subj ects are

included in the simulation. The 75 subj ects' parameters were assumed to follow the same

distribution as the 14 subj ects' whose population means and variances were estimated in

above analysis. Each subj ect' s parameters were randomly sampled from the estimated

populations respectively, and kept the same throughout the simulation. For those

parameters whose inter-subj ect variability was not estimated, the values of population

mean, either estimated or fixed in above analysis, were applied to each subj ect in the

simulation. The whole simulation for the four dosing regimens was repeated 8 times with

different seed numbers assigned in NONMEM, which were to generate different groups

of 75 hypothetic subj ects in each simulation. For each simulation, the 75 subj ects' 24-

hour CCS at steady state resulted from different administration time were calculated and

then compared, respectively for each dosing regimen, using Friedman test followed by

Dunn's multiple comparison. Similarly, the 75 subjects' 24-hour CLS at steady state

resulted from different administration time were calculated and then compared,

respectively for each dosing regimen in each simulation, using Friedman test followed by

Dunn's multiple comparison.

Results

Population Pharmacokinetic Model

Budesonide

A one-compartment model was chosen as structure model for BUD since two-

compartment model was over-parameterized and one-compartment model was sufficient

to describe the data (Appendix A for NONMEM code). No trend was observed in the

diagnostic plot of weighted residuals versus time, indicating no bias in the model













































6n ~ .Pa ~'~9,
;aa 8 B a
a a
~~"a, B s M ~
a a a a ~ib
a a
O0

0 20 40 60 80 100 120
Tim e (h r)






$i''" si
s '~'~'


prediction. The relationship between population predicted, individual predicted and


observed plasma concentrations as a function of time is shown in Figure 5-1. There was


no significant inter-occasion variability found for any parameter (Table 5-1) based on


likelihood ratio test (P<0.01, X20.05 = 6.64, df = 1). The parameter estimates were in good


agreement with the mean parameter estimates of the bootstrap replicates (Table 5-1),


indicating stability of the model [143]. The value for the absorption rate (Ka) was fixed


because BUD's fast absorption and plasma sampling time scheme did not allow


estimation of this parameter.


0 20 40 60
Time (hr)


80 100 120


Figure 5-1. Relationship between population predicted (open triangle) and observed BUD
(open circle) concentrations, individual predicted (open square) and observed BUD
(open circle) concentrations as function of time.









Table 5-1. Summary of BUD pharmacokinetic model parameter estimates
Parameters Original dataset 200 bootstrp replicates
estimate (RSE %) mean (RSE %)
6Ka, hl 10 (Fixed) 10 (Fixed)
6K10, h- 0.21 (5.5) 0.21 (5.3)
8v, L 1320 (9.8) 1336 (9.2)
UOKa, %CV
UK10, %CV 15.5 (85.4) 14.2 (46.5)
coy, %CV 31.2 (37.2) 30.0 (20.6)
o, ng/ml 0.54 (17.5) 0.54 (8.7)
RES, relative standard error; CV, coefficient of variation; Ka, absorption rate; K10,
elimination rate constant; V, volume of distribution central compartment; = not
calculated.


Fluticasone propionate

A two-compartment model was chosen as structure model for FP since one-

compartment model did not fit data well (Appendix B for NONMEM code). No trend

was observed in the diagnostic plot of weighted residuals versus time. The relationship

between population predicted, individual predicted and observed plasma concentrations

as a function of time is shown in Figure 5-2. There was no significant inter-occasion

variability found on any parameter (Table 5-2) based on likelihood ratio test (P<0.01,

X20.05 = 6.64t, df = 1). The parameter estimates were in good agreement with the mean

parameter estimates of the bootstrap replicates (Table 5-2), indicating stability of the

model .

Population Pharmacokinetic/Pharmacodynamic Model

The population PK/PD model provided reasonable fits for the cortisol and

lymphocyte data (Appendix C and D for NONMEM code). No trend was observed in the

diagnostic plot of weighted residuals versus time. The relationship between population

predicted, individual predicted and observed plasma concentrations as a function of time

is shown in Figure 5-3, and 5-4. The parameter estimates are listed in Table 5-3. A






























O 6 ph

A d B C~
II Y
B a~ ~
.d W
Y1 k:
n ~oO~
a o r-l II


significant inter-occasion variability was found for Rmax for both BUD and FP. None of


the 24-hour CCS and CLS at steady state calculated from simulated cortisol


concentrations and lymphocyte counts time-profile was different from observed CCS and


CLS in the original study [1 12], respectively (p < 0.05), indicating stability of the model.


~ -2.5

of
ti~
o_-rn
e, ---t.v
o
a
o
na,
or
am

" 95.5









of

~$4.0
o
a,
ulo
mo
90
-5.5


0 20 40 60
Time (hr)


80 100 120















80 100 120


0 20 40 60
Time (hr)


Figure 5-2. Relationship between population predicted (open triangle) and observed FP
(open circle) concentrations, individual predicted (open square) and observed
FP (open circle) concentrations as function of time.



Population Pharmacokinetic/Pharmacodynamic Simulation


The relationship between administration time and corresponding mean CCS and


CLS of the 75 subj ects from each simulation is shown in Figure 5-5 and 5-6 (Appendix E


and F for NONMEM code). For once-daily of BUD, maximum CCS was observed when


BUD was given in 2am-5am, while minimum CCS was observed when BUD was given


I









in 3pm-6pm. Similarly, for once-daily of FP, maximum CCS was observed when FP was

given in 4am-8am, and minimum CCS was observed when FP was given in 3pm-8pm.

On the contrary to what we observed for CCS, maximum CLS was observed when BUD

was given in 3pm-8pm and when FP was given in 3pm-7pm, while minimum CLS was

observed when BUD or FP was given in 3am-9am. The maximum and minimum mean

CCS and CLS of the subj ects and the corresponding administration time points of the 8

simulations are summarized in Table 5-4 and 5-5. The difference between the maximum

and minimum mean suppression of the subj ects are also summarized in Table 5-4 and 5-5.

The maximum mean CCS of the subj ects is significantly different from minimum mean

CCS of the subj ects, respectively for each dosing regimen of each simulation (p < 0.001).

In addition, the maximum mean CLS of the subj ects is significantly different from

minimum mean CLS of the subj ects, respectively for each dosing regimen of each

simulation (p < 0.001).

Table 5-2. Summary of FP pharmacokinetic model parameter estimates
Parameters Original dataset 200 bootstrp replicates
estimate (RSE %) mean (RSE %)
6Ka, hl 2.79 (13.7) 3.3 (39.9)
6Kl2, h- 0.21 (13.0) 0.24 (56.7)
6K21, h- 0.16 (8.0) 0.18 (26.6)
6K10, h- 0.14 (6.9) 0.13 (11.7)
8v, L 4130 (8.0) 4454 (18.2)
UOKa, %CV 13.7 (95.2) 11.4 (60.8)
UKl2, %CV 16.1 (100.8) 18.6 (72.4)
UK21, %CV 30.1 (67.1) 28.5 (77.9)
UK10, %CV 14.5 (70.3) 14.4 (96.3)
coy, %CV 10.7 (80.6) 9.2 (49.6)
o, ng/ml 0.27 (10.0) 0.26 (5.5)
RES, relative standard error; CV, coefficient of variation; Ka, absorption rate; Kl2,
transfer rate constant from central compartment to peripheral compartment; K21,
transfer rate constant from peripheral compartment to central compartment ; K10,
elimination rate constant; V, volume of distribution central compartment; = not
calculated.




















2200



-
01



--


10


0 20 40 60
Time (hr)


80 100 120


0 20 40 60 80 100 120
Time (hr)


Figure 5-3. Diagnostic plots of BUD PK/PD model. (a) Relationship between individual
predicted (open square) and observed (open circle) cortisol concentrations as
function of time for BUD PK/PD model. (b) Relationship between individual

predicted (open square) and observed (open circle) lymphocyte counts as
function of time for BUD PK/PD model.


o rr
oo 8















L 8 80
I~
B o r o~i


B I '


I a o OR


0 20 40 60 80 100 120
Tln~ (hr)


oYMa o
h~e~o~8E




~a~"~


200



8100



0





130





--

~8 so


0 20 40 60
Tlrne (hr)


80 100 120


Figure 5-4. Diagnostic plots of FP PK/PD model. (a) Relationship between individual
predicted (open square) and observed (open circle) cortisol concentrations as function of
time for FP PK/PD model. (b) Relationship between individual predicted (open square)
and observed (open circle) lymphocyte counts as function of time for FP PK/PD model.



































9.4 (42.9)
17.6 (44.7)
60.0 (47.1)
n/c
7.8 (162.6)
35.6 (69.0)
80.4 (36.6)
0.54 (8.7)
0.17 (12.2)


FP


Parameters BUD

Mean (RSE %)


I \


Table 5-3. Summary of FP and BUD pharmacokinetic/pharmacodynamic model
parameter estimates


22.7 (2.5)
17.5 (0.7)
33.7 (Fixed)
0.56 (Fixed)
3270 (4.1)
0.02 (20.9)
21.7 (25.3)
0.19 (28.4)
30.8 (15.6)
0.16 (36.9)
7.0 (54.1)
4.0 (33.9)


Mean (RSE %/)
22.7 (2.5)
17.5 (0.7)
33.7 (Fixed)
0.56 (Fixed)
2490 (7.6)
0.008 (37.7)
14.4 (60.8)
0.13 (77.5)
26.8 (31.8)
0.05 (89.1)
7.0 (54.1)
4.0 (33.9)


21.8 (54.9)
15.4 (59.3)
46.0 (158.0)
n/c
6.3 (127.9)
71.7 (80.4)
88.7 (133.3)
0.54 (10.0)
0.17 (15.2)


9Tmax, h
9Tmni, h
91 d, L
9Ke, h'
Rma,, ng/h
9EC50ICSaco*, ng/ml
9Knz, %//h
9Kout, %//h
BEC5(1Cortalym, ng/ml
9EC50ICSalym, ng/ml
O)Tmax, %CV
O)Tmni, %CV
01Vd.%CV
03Ke, %CV
o~ma, H1-, %CV
ORmax, 101-, %CV
O)EC5(1ICSacoI*, %CV
)Knz, %CV
U)Kout, %CV
O)EC5(1Cortalym, %CV
O)EC5(1ICSalymf %CV
Geortisol, ng/ml
Ol mnhnroo res.%







71







30 ~30
S20 i ii uuuuuuuuuuuuuuuuuuuu S20

0 0



Administration Time Adnumstration Time


c dC




S15 ~15





Adnuntstratton Time Adnumstration Time


Figure 5-5. Relationship between administration time and corresponding mean (+ SE)
cumulative suppression. a) CCS of the 75 subjects after once-daily 1mg of
BUD. b) CCS of the 75 subjects after once-daily 2mg of BUD. c) CLS of the
75 subjects after once-daily 1mg of BUD. d) CLS of the 75 subjects after
once-daily 2mg of BUD. Empty circle: simulation 1; Filled circle: simulation
2; Empty triangle: simulation 3; Filled triangle: simulation 4; Empty Square:
simulation 5; Filled Square: simulation 6; Empty heart: simulation 7; Filled
heart: simulation 8.

















b.


Administration Time


Adnm~stmt~on Tune


d


c


Adnm~stat~on Tune


Adnm~staton Tune


Figure 5-6. Relationship between administration time and corresponding mean (+ SE)
cumulative suppression. a) CCS of the 75 subjects after once-daily 0.5mg of
FP. b) CCS of the 75 subjects after once-daily 1mg of FP. c) CLS of the 75
subjects after once-daily 0.5mg of FP. d) CLS of the 75 subjects after once-
daily 1mg of FP. Empty circle: simulation 1; Filled circle: simulation 2;
Empty triangle: simulation 3; Filled triangle: simulation 4; Empty Square:
simulation 5; Filled Square: simulation 6; Empty heart: simulation 7; Filled
heart: simulation 8.











Table 5-4. Summary of the mean cumulative cortisol suppression of the 75 subjects of the
8 simulations
Corticosteroid Maximum & SD" Administration Minimum & SDe Administration F~luctuationd
(%) time point (%) time point" 04)
BUD 1 mg 27.5 & 0.9 2am 5am -3.5 A 1.9 3pm 6pm 31.0 & 2.5
2 mg 44.0 & 1.6 3am- 5am 12.5 A 1.2 4pm 5pm 31.5 & 2.0
FP 0.5 mg 12.9 & 2.6 4am 8am -4.4 & 2.4 5pm 8pm 17.3 & 3.3
1 mg 22.1 + 1.3 4am 5am -0.2 & 1.4 3pm 6pm 22.3 & 3.1
a. The mean and standard deviation of the maximum mean CCS of the 75 subjects of the
8 simulations.
b. The range of administration time points which corresponds to the maximum mean CCS
of the 75 subjects of the 8 simulations.
c. The mean and standard deviation of the minimum mean CCS of the 75 subj ects of the
8 simulations.
d. The range of administration time points which corresponds to the minimum mean CCS
of the 75 subjects of the 8 simulations.
e. The mean and standard deviation of the difference between maximum mean CCS and
the minimum mean CCS of the 75 subjects of the 8 simulations.


Table 5-5. Summary of the mean cumulative lymphocytes suppression of the 75 subjects
of the 8 simulations
Corticosteroid Maximum & SDa Administration Minimum & SDe Administration Fluctuationd
(%) time point (%) time point" 04)
BUD 1 mg 9.6 & 0.5 4pm 8pm 4.7 & 0.6 5am 9am 4.9 & 0.6
2 mg 15.8 & 0.7 3pm 8pm 9.8 & 1.0 3am 5am 6.0 & 1.5
FP 0.5 mg 6.8 & 0.3 4pm 7pm 2.5 & 0.4 3am 8am 4.2 & 0.5
1 mg 9.6 & 0.4 3pm 6pm 4.5 & 0.4 5am 9am 5.2 & 0.4
a. The mean and standard deviation of the maximum mean CLS of the 75 subjects of the
8 simulations.
b. The range of administration time points which corresponds to the maximum mean CLS
of the 75 subjects of the 8 simulations.
c. The mean and standard deviation of the minimum mean CLS of the 75 subjects of the 8
simulations.
d. The range of administration time points which corresponds to the minimum mean CLS
of the 75 subjects of the 8 simulations.
e. The mean and standard deviation of the difference between maximum mean CLS and
the minimum mean CLS of the 75 subj ects of the 8 simulations.


Discussion

Without exogenous stimuli, both endogenous cortisol concentrations and

lymphocyte counts undergo a circadian rhythm. Moreover, the circadian rhythm of

lymphocyte counts is inversely correlated with the circadian rhythm of endogenous









cortisol [121, 144]. One therefore should expect that administration time of an ICS has an

effect on not only the cortisol but also lymphocytes suppression. In the current study we

employed a population PK/PD simulation approach to evaluate the effect of the

administration time on cumulative cortisol suppression and cumulative lymphocytes

suppression simultaneously, and to identify an optimum administration time point for

once-a-day dosing regimen in terms of minimum CCS and CLS, respectively.

A population PK/PD model was developed in this study to describe the

disposition of BUD and FP after inhalation, the effect of BUD and FP on cortisol, and the

combined effect of BUD or FP and cortisol on lymphocytes in healthy subj ects. The

structure (PK/PD) model was a mechanism-based model and previous study has

demonstrated the validity of the model via standard two-stage analysis [121]. The current

study successfully applied the model to non-linear mixed effects modeling analysis and

estimated the inter and intra-subj ect and inter-occasion variability of model. After

repeated inhalation, BUD plasma concentrations time-profiles were well described by

one-compartment model with first-order absorption, while FP plasma concentrations

time-profile was well described by two-compartment model with first-order absorption.

The extra disposition of FP to a peripheral compartment compared with BUD could result

from the more sensitive assay, about 5 folds more sensitive than assay for BUD, and its

higher lipophilicity than BUD's. The PK parameters of BUD and FP derived from this

study are in broad agreement with previously reported values [145]. The steroid-

independent pharmacodynamic parameters (Tmax, Tmin, Rmax, Kin and Kout),

EC50IC S~cort of BUD and FP, EC50IC S41ym of BUD and EC50Cortalym observed

with the non-linear mixed effect modeling approach are comparable to results of previous










studies [75, 121, 146]. A linear relationship between EC50 for cortisol suppression for

various corticosteroids and their relative receptor affinity (RRA), estimated in vitro, has

been established in previous studies [147]. Hence, a linear relationship between EC50 for

lymphocytes suppression for various corticosteroids and their RRA is reasonably

expected. If we plot the 1/RRA versus EC50 of BUD, cortisol and FP for suppression on

lymphocytes obtained in this study, a good linear curve (R2 = 0.95) could be established

and should prove useful in future prediction of EC50 of other steroids (in vivo effects)

based on their in vitro RRA (Figure 5-7).


1/EC50 = -1.373 + 0.01109*RRA


F P*


18



S13



8-
I / BUD



Cortisol





50 300 550 800 1050 1300 1550 1800
RRA (Relative Receptor Affinity)


Figure 5-7. Relationship between ECso, free of steroids for lymphocytes suppression and
their reciprocal of the relative receptor affinity.









For both drugs, the maximum CCS was observed when ICS was administered in

the early morning and the minimum CCS was observed when ICS was administered in

the afternoon, and this observation is in agreement with previous report [127]. This

circadian rhythm results from the interaction between the systemic activity of exogenous

corticosteroids and endogenous cortisol release: CCS is maximized when the period of

high systemic activity of exogenous corticosteroids (concentrations higher than EC50)

covers the period of high cortisol release in the early morning; CCS is minimized when

the period of high systemic activity covers the period of low cortisol release in the late

night. On the contrary, the maximum CLS was observed when ICS was given in the

afternoon while the minimum CLS was observed when ICS was given in the early

morning. This circadian rhythm is the result of combined modulation of natural circadian

rhythm of the lymphocytes release by systemic activity of exogenous corticosteroids and

endogenous cortisol, whose release is suppressed by the same systemic exogenous

corticosteroids. Hence, the CLS reaches the maximum when the period of high systemic

activity of the net corticosteroids, the total of the suppressed endogenous cortisol and

exogenous corticosteroids, covers the period of high lymphocytes release in the late night.

And it reaches the minimum when the period of high systemic activity of the net

corticosteroids covers the period of low lymphocytes release in the early morning. So for

once-a-day dosing regimen, the optimum administration time is in the afternoon in terms

of minimized CCS, however, the optimum administration time is in the morning in terms

of minimized CLS.

In our simulations, the administration time for which maximum suppression of

cortisol and blood lymphocytes was observed differed. The advantage to use cortisol










suppression as biomarker is that cortisol is more sensitive to the presence of exogenous

corticosteroids since the EC50 values for the effect of ICS on lymphocytes was about 1 -

8 times higher than that for the effect on cortisol as demonstrated in previous and this

study [146, 148]. It is, however, disputed whether cortisol suppression is a good marker

for systemic effects or side effects, as the mechanism of cortisol suppression is likely to

not reflect anti-inflammatory events or effects on growth that is affected by both

endogenous and exogenous glucocorticoids.

While the lymphocytes suppression is somewhat less sensitive than the serum

cortisol suppression, the immunological character of lymphocytopenia, has the advantage

of representing a more relevant pharmacodynamic endpoint, that is affected by both

endogenous and exogenous glucocorticoids. Lymphocytopenia is a systemic effect, and

should represent a side effect for inhalation therapy. According to this definition delivery

of ICS in the evening hours would induce the highest degree of side effects. Similar,

administration of a systemic glucocorticoid (e.g. oral delivery) would induce the

maximum effects when given in the evening. On the other hand, it is also likely that the

desired pulmonary effects of inhaled glucocorticoids are the sum of endogenous and

exogenous glucocorticoids. Consequently, one would also predict maximum

antiasthmatic effects for an evening administration. The differences observed in the

simulations between the two biomarkers supports the note that these biomarkers are not

interchangeable on a qualitative and quantitative level.

The terminal half-life of FP after inhalation is much longer than that of BUD (14.3

hours vs. 3.3 hours as estimated from this study). It has been recognized that for a

receptor mediated response, the duration and decline of this response depends partly on









the elimination rate of the drug [149, 150]. For FP, a longer terminal life leads to less

fluctuated concentration during the dosing interval fluctuate less. This resulted in our

"dosing time" simulations in a smaller difference between maximum and minimum

cumulative cortisol suppression for 500 and 1000 Clg FP than for 1,000 and 2,000 Clg

BUD (Table 5-4). Thus, the dosing time is less important for the CCS of FP. Interestingly,

the difference between maximum and minimum CLS after once-daily administration of

1mg or 2mg BUD and after once-daily of 0.5 or 1mg FP were rather similar (Table 5-4),

which is likely be related to the rather small overall effect on blood lymphocytes and the

bufferingg" capacity of cortisol.

Once-a-day dosing has been successfully evaluated for some of the newer inhaled

glucocorticoids (budesonide, ciclesonide) because of the easy compliance [48]. Our

results indicate that the administration time may be important for once-a-day dosing

regimen and that a late afternoon or evening administration is likely to result in maximum

effects. Clinical studies evaluating the effect of time of dosing on ICS efficacy and side

effects are limited. Some clinical studies suggested that afternoon or early evening dosing

tends to be superior in efficacy to a morning dosing especially for patients with nocturnal

asthma [151, 152]. This is in agreement with our simulation results when data on

systemic lymphocytopenia are applied to the local pulmonary effects.

Conclusion

The population PK/PD modeling/simulation presented in current study is a step

further into predicting outcome of clinical study based on previous developed PK/PD

model. The population approach taken in our study incorporates both inter-subj ect and

intra-subj ect variability, and allows probing for significant differences. Previous non-

population PK/PD simulation approaches have already proven high predictive power for










CCS [153], our population PK/PD simulation should be even more informative with

respect to the statistical evaluation of the simulated events. Employing this approach, we

demonstrated the administration time is an important factor to minimize the systemic side

effects and maximize the benefit/risk ratio for once-daily regimen of ICS. However,

further clinical studies are needed to verify the prediction by our simulation.














CHAPTER 6
CONCLUSIONS

Pulmonary targeting is a desired characteristic for successful inhaled

corticosteroids. The overall goal of this study was to evaluate certain aspects of factors

affecting pulmonary targeting of current ICSs, hence to gain more understanding of those

factors and contribute to the journey of developing "ideal" ICS for asthmatic patients.

Our first study investigated how different valved holding chambers (VHCs) affect

fluticasone propionate aerosols delivered from a metered dose inhaler (MDI). It was

found that all three VHCs tested, AeroChamber-PlusTM"', OptiChambers and

OptiChambero-Advantage significantly reduced the in vitro oropharyngeal deposition of

fluticasone propionate. It was also found that the respirable dose from the MDI attached

to AeroChamberTM-Plus or OptiChambero was not significantly different from the

respirable dose from fluticasone propionate MDI alone, while the respirable dose of

fluticasone from OptiChambero'-Advantage was 40-45% less than from the MDI alone.

This result suggests that AeroChamberTM-Plus or OptiChambers could be successfully

used with fluticasone MDI since both VHCs not only could reduce the risk of topical

adverse effects induced by fluticasone, but also deliver the same amount of respirable

dose of fluticasone, while OptiChambero-Advantage would not be a good choice for

patients on fluticasone MDI since much less of respirable dose of fluticasone was

delivered by it. This study shows that choice of VHC is an important factor for

determination of pulmonary targeting of fluticasone propionate.









Our second study showed that the systemic absorption of beclometasone

monopropionate, budesonide, fluticasone propionate and mometasone furoate appears to

be greater in patients with a higher FEV1 % predicted although other factors, including

age and gender affect this relationship. In the third study, using a methacholine challenge

we were able to study the effect of change in lung function within an individual on

plasma concentrations of fluticasone and budesonide under controlled conditions. A 33 %

reduction in FEV1 in our study caused a 60 % reduction in AUCO-5h ValUeS for fluticasone

and 29% reduction in AUCo-s values for budesonide, suggesting a 60 % reduction and a

29% in systemic exposure for fluticasone and budesonide respectively. Results from

these two studies indicate that the systemic exposure, hence the risk adverse systemic

effects from ICSs is correlated with asthmatic patient' s lung function, and extent of the

correlation varies for different ICSs. We suggest that a difference in lipophilicity among

ICSs probably is the reason for the difference in the extent of correlation among various

ICSs. Further study is necessary to explore this. Our Eindings re-enforce the importance

of reviewing the need for higher doses of inhaled corticosteroids and particularly

fluticasone, as lung function improves and in patients with relatively normal lung

function, to improve the benefit to risk ratio of ICSs.

Use of biomarker, such as cortisol suppression and lymphocytes facilitates the

evaluation of systemic effects of different ICS regimens. Population PK/PD modeling

and simulation provides us a great tool to describe the relationship between drug

concentrations and biomarker concentrations, and to predict and assess statistical

significance of the outcomes of future studies. Hence, we developed a population PK/PD

model for inhaled budesonide and fluticasone, respectively to describe the relationship









between drug concentrations, cortisol concentrations and lymphocyte counts. Further, we

used this model to evaluate the administration time effect on the cumulative cortisol

suppression (CCS) and lymphocytes suppression (CLS) for once-daily inhaled BUD and

FP, respectively. Our results indicates that the optimal time for administration of BUD or

FP, in terms of minimized systemic side effects, depends on the choice of the biomarker:

the minimum CCS is achieved when ICS is given in the afternoon, while minimum CLS

is achieved when ICS is given in the morning. This administration time effect is the result

of interaction between natural circadian rhythm of endogenous cortisol concentration and

lymphocyte concentrations, and the effect of PK/PD properties of ICSs on these two

rhythms. Our study also showed that both the difference between maximum CCS and

minimum CCS, and the difference between maximum CLS and minimum CLS were

significant (P <0.001). The extent of difference between maximum and minimum CCS

suppression differs between BUD and FP because their different PK properties. However,

whether the difference between maximum and minimum suppression is clinically

significant needs further clinical studies to evaluate.
















APPENDIX A
NONMEM CODE FOR PK MODEL OF BUD AFTER INHALATION

$PROBLEM BUD PK MODEL
$INPUT ID TIME AMT DV CMT EVID MDV OCC
$DATA LNPK1. CSV IGNORE= #
$SUBROUTINES ADVAN2 TRANS1
$PK
TVKA= THETA (1) ; ABSORPTION
TVK= THETA (2)
TVVC=THETA(3)
KA=TVKA
K=TVK+EXP(ETA(1))
VC=TVVC+EXP(ETA(2))
CL=K+VC
S2=VC
$ERROR
IPRED=0
IF (F. GT. 0) IPRED= LOG (F)
Y= IPRED+ERR (1)
$THETA
(10 FIXED)
(0 0. 22)
(0 1080)
$OMEGA 0. 1 0. 1
$SIGMA 0. 04
$ESTIMATION METHOD=1 NOABORT MAXEVAL=9999 PRINT=5 POSTHOC
$COV PRINT=E
$TABLE ID TIME DV IPRED KA K VC CL ETA(1) ETA(2)
NOPRINT ONEHEADER FILE=BUDPK. FIT
















APPENDIX B
NONMEM CODE FOR PK MODEL OF FP AFTER INHALATION

$PROBLEM FP PK MODEL
$INPUT ID TIME AMT DV CMT EVID MDV OCC
$DATA LNFPPK. CSV IGNORE=
$SUBROUTINES ADVAN4 TRANS1
$PK
TVKA= THETA (1) ; ABSORPTION
TVK= THETA (2)
TVK23=THETA(3)
TVK32=THETA(4)
TVVC=THETA(5)
KA= TVKA+EXP (ETA (1))
K=TVK+EXP(ETA(2))
K23=TVK23+EXP(ETA(3))
K32= TVK32+EXP (ETA(4) )
VC= TVVC+EXP (ETA (5))
CL=K+VC
S2=VC
$ERROR
IPRED=0
IF (F. GT. 0) IPRED= LOG (F)
Y= IPRED+ERR (1)
$THETA
(0 4)
(O 0. 1)
(0 0. 159)
(O 0. 1)
(0 2000)
$OMEGA 0. 1 0. 15 0. 1 0. 1 0. 1
$SIGMA 0. 1
$ESTIMATION METHOD=0 NOABORT MAXEVAL=9999 PRINT=5 POSTHOC
$COV PRINT=E
$TABLE ID TIME DV IPRED KA K K23 K32 VC CL ETA(1) ETA(2)
ETA(3) ETA(4) ETA(5) NOPRINT ONEHEADER FILE=FPPK. FIT
















APPENDIX C
NONMEM CODE FOR PK/PD MODEL OF BUD

$PROBLEM CORTISOL BASELINE PD MODEL
$INPUT ID TIME AMT DV EVID CMT MDV
$DATA Lncortbase.CSV IGNORE=#
$SUBROUTINES ADVAN6 TRANS1 TOL=5
$MODEL
COMP=(CORTISOL);1
$ABBREVIATED DERIV2=NO
$PK
TVTMAX=THETA(1); TIME OF MAX RELEASE
TVTMIN=THETA(2); TIME OF MINIMUM RELEASE
TVRMAX=THETA(3); MAXIMUM RELEASE RATE
VD=33.7; DIST VOLUME OF CORTISOL
KE=0.56; ELIMINATION RATE OF CORTISOL
TMAX=TVTMAX+EXP(ETA(1))
TMIN=TVTMINSEXP(ETA(2))
RMAX=TVRMAX+EXP(ETA(3))
S1=VD
Rel=RMAX/(TMAX-TMIN-24)*TIME-RMAX+TMIN/(TMAXTI24
Rc2=RMAX/(TMAX-TMIN)*TIME-RMAX+TMIN/(TMAX-TMN
Rc3=RMAX/(TMAX-TMIN-2~TIE 4)*(AXTMIME2)RAXTI/TAXTI4
Rc4=RMAX/(TMAX-TMIN)*(TIME-24)-RMAX+TMIN/(TMA-MN
Re5=RMAX/(TMAX-TMIN-2~(IM 4 )*(AXTMIME4)RAXTI/TAXTI4
Rc6=RMAX/(TMAX-TMIN)*(TIME-48)-RMAX+TMIN/(TMA-MN
Rc7=RMAX/(TMAX-TMIN-24)*(ME2RMXTMIME7)RAXTI/TAXTI4
Rc8=RMAX/(TMAX-TMIN)*TM7)RA~MN(TIM-2-MXTI/TA-MN
Re9=RMAX/(TMAX-TMIN-24)*(TIME96)-RAX+MIN(TAX-MIN24
Rel0=RMAX/(TMAX-TMIN)*(ME6RMXTMIME9)RMXTIN(MXMN
Roll=RMAX/(TMAX-TMIN-24)*(TIME-120)-RMAX+TMI N(AXTI24
Rel2=RMAX/(TMAX-TMIN>(IE2)*(AXTMIME10-A+MN(TA-MN
Rel3=RMAX/(TMAX-TMIN-2~(IM1 4)*(XTMIM14)RATIN(AXMN2)
Rel4=RMAX/(TMAX-TMIN)*(TIME-144)-RMAX+TMIN/ (TA-MN
M1=0
IF (TMIN.GT.TIME) M1=1
M2=0
IF (TIME.GT.TMAX) M2=1
M3=0
IF (TMIN+24.GT.TIME) M3=1
M4=0
IF (TIME.GT.TMAX+24) M4=1










M5=0
IF (TMIN+48.GT.TIME) M5=1
M6=0
IF (TIME.GT.TMAX+48) M6=1
M7=0
IF (TMIN+72.GT.TIME) M7=1
M8=0
IF (TIME.GT.TMAX+72) M8=1
M9=0
IF (TMIN+96.GT.TIME) M9=1
M10=0
IF (TIME.GT.TMAX+96) M10=1
M11=0
IF (TMIN+120.GT.TIME) M11=1
M12=0
IF (TIME.GT.TMAX+120) M12=1
M13=0
IF (TMIN+144.GT.TIME) M13=1
M14=0
IF (TIME.GT.TMAX+144) M14=1
Reortl=Rel*M1+Rc2+(1-M1)*(1-)+cM2)+R3+M2M+c+1M)(-4
Reort2=Re5+M4+M5+Rc6+(1-M5)*(1M6)+Rc7+M6+M+c81M)(-8
Reort3=Re95M88M9+Rcl0+(1-M9)*(1M10)+Rell*M0M1Rl+1M1*1M2
Reort4=Rel3+M12+M13+Rel4+(1-M13)*(1-M14
Rcort=Rcortl+Rcort2+Reort3+Reort4
$DES
DADT(1)=Reort-KEEA(1)
$ERROR
IPRED=0
IF(F.GT.0) IPRED=LOG(F)
Y=IPRED+ERR(1)
$THETA
(0 22 24)
(0 18 24)
(0 5000 20000)
$OMEGA
(0. 2)
(0. 2)
(0. 2)
$SIGMA
(0.1I)
$ESTIMATION METHOD=0 NOABORT MAXEVAL=9999 PRINT=5 POSTHOC
$COV PRINT=E
$TABLE ID TIME DV IPRED TMAX TMIN RMAX ETA(1) ETA(2)
ETA(3) NOPRINT ONEHEADER FILE=CORTBASE. FIT










$PROBLEM BUD CORTISOL PK/PD MODEL
$INPUT id time amt dv evid amt mdv occ ika ik ive itma itmi irma
$DATA Lncort.CSV IGNORE=#
$SUBROUTINES ADVAN6 TRANS1 TOL=5
$MODEL
COMP=(DEPOT, DEFDOSE) ;1
COMP=(CENTRAL) ;2
COMP=(CORTISOL);3
$ABBREVIATED DERIV2=NO
$PK
KA=IKA
K=IK
VC=IVC
S2=VC
TMAX=ITMA; TIME OF MAX RELEASE
TMIN=ITMI; TIME OF MINIMUM RELEASE
BOV1=ETA(2)
IF (0CC.EQ.2) THEN
BOV1=ETA(3)
ENDIF
BSV1=ETA(1)
BOV2=ETA(5)
IF (0CC.EQ.2) THEN
BOV2=ETA(6)
ENDIF
BSV2=ETA(4)
RMAX=THETA(1)*EXP(BSV1+BOV1); MAXIMUM RELEASE RATE
VD=33.7; DIST VOLUME OF CORTISOL
KE=0.56; ELIMINATION RATE OF CORTISOL
S3=VD
EC50=THETA(2)*EXP(BSV2+BOV2)

Rel=RMAX/(TMAX-TMIN-24)*TIME-RMAX+TMIN/(TMAXTI24
Rc2=RMAX/(TMAX-TMIN)*TIME-RMAX+TMIN/(TMAX-TMN
Rc3=RMAX/(TMAX-TMIN-2~TIE 4 )*(AXTMIME2)RAXTI/TAXTI 4
Rc4=RMAX/(TMAX-TMIN)*(TIME-24)-RMAX+TMIN/(TMA-MN
Re5=RMAX/(TMAX-TMIN-2~(IM 4 )*(AXTMIME4)RAXTI/TAXTI 4
Rc6=RMAX/(TMAX-TMIN)*(TIME-48)-RMAX+TMIN/(TMA-MN
Rc7=RMAX/(TMAX-TMIN-24)*(ME2RMXTMIME7)RAXTI/TAXTI 4
Rc8=RMAX/(TMAX-TMIN)*TM7)RA~MN(TIM-2-MXTI/TA-MN
Re9=RMAX/(TMAX-TMIN-24)*(TIME96)-RAX+MIN(TAX-MIN24
Rel0=RMAX/(TMAX -TMIN)*(ME6RMXTMIME9)RMXTIN(MXMN
Roll=RMAX/( TMAX -TMIN-24)*(TIME-120)-RMAX+TMIN(AXTI24
Rel2=RMAX/(TMAX -TMIN>(IE2 )*(AXTMIME10-A+MN(TA-MN
Rel3=RMAX/( TMAX -TMIN-2~(IM1 4 )*(XTMIM14)RATIN(AXMN2)
Rel4=RMAX/( TMAX -TMIN)*(TIME-144)-RMAX+TMIN/(TA-MN










M1=0
IF (TMIN. GT. TIME) M1=1
M2=0
IF (TIME. GT. TMAX) M2=1
M3=0
IF (TMIN+24. GT.TIME) M3=1
M4=0
IF (TIME. GT. TMAX+24) M4=1
M5=0
IF (TMIN+48. GT.TIME) M5=1
M6=0
IF (TIME. GT. TMAX+48) M6=1
M7=0
IF (TMIN+72. GT.TIME) M7=1
M8=0
IF (TIME. GT. TMAX+72) M8=1
M9= 0
IF (TMIN+96. GT.TIME) M9=1
M10=0
IF (TIME. GT. TMAX+96) M10=1
M11=0
IF (TMIN+120. GT. TIME) M11=1
M12=0
IF (TIME. GT. TMAX+120) M12=1
M13=0
IF (TMIN+144. GT. TIME) M13=1
M14=0
IF (TIME. GT. TMAX+144) M14=1
Reor tl= Rel*M1+Rc2+ (1-M1) *(1-)+cM2) Rc3+MM3R4(13)*1M)
Reort2=Re5+M4+M5+Rc6+(1-M5)*(1M6)+Rc7+M6+M+c81M)(-8
Reort3=Re95M88M9+Rcl0+(1-M9)*(1M10)+Rell*M0M1Rl+1M1*1M2
Reort4=Rel3+M12+M13+Rel4+(1-)(M13)(1M4
Rcort=Rcortl+Rcort2+Reort3+Reort4
$DES
DADT (1)=KASA (1)
DADT (2) =KASA(1) KSA(2)
CP=A(2)/S2
EFF=1-CP/(EC50+CP)
DADT (3)=Reort*FEFF-KESA (3)
CORT=A (3) /S3
$ERROR
IPRED=0
IF (F. GT. 0) IPRED= LOG (F)
Y= IPRED+ERR (1)
$THETA
(0 3100 20000)
(0 0. 2)




Full Text

PAGE 1

FACTORS AFFECTING PULMO NARY TARGETING OF INHALED CORTICOSTEROIDS By KAI WU 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 2006

PAGE 2

Copyright 2006 by Kai Wu

PAGE 3

For my Waipo, Baba, Mama, Gege and Laopo.

PAGE 4

ACKNOWLEDGMENTS I extend my appreciation and grateful thanks to Dr. Guenther Hochhaus for guiding me through the four years of Ph.D. study, giving me excellent suggestions and always supporting me. I would like to thank the members of my supervisory committee, Dr. Hartmut Derendorf, Dr. Jeffrey Hughes, and Dr. Saeed R. Khan, for their valuable and kind advice throughout my research. I take this opportunity to express my gratitude to Yufei Tang for her invaluable technical assistance. I would like to thank all the secretaries, Jame Ketcham, Andrea Tucker and Patricia Khan for their technical and administrative assistance. I thank my group members, Yaning, Manish, Intira, Vikram, Sriks, Zia, Elanor, Keerti, Navin, Nasha and other graduate students and post-docs in the department for their assistance and friendship. Especially I want to thank all my friends, Weihui, Chao, Jeff, Victor, Aixin, Kai and Zhen, Chengguan and Qin, Yipeng and Hongxiao and all my other friends for their help and support that have made my stay in Gainesville a pleasant experience. I want to express my deep gratitude and love for my parents, brother and other family members for their unconditional support and encouragement of my academic pursuits over the years. Last, but not least, I want to thank my lovely wife, Ling, for her 120% support, full understanding and timely encouragement. Without her, I would not have been able to achieve any of what I have achieved. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES ...........................................................................................................viii LIST OF FIGURES ...........................................................................................................ix ABSTRACT .......................................................................................................................xi CHAPTER 1 INTRODUCTION........................................................................................................1 Asthma..........................................................................................................................3 Mechanism of Action of Corticosteroids in Asthma....................................................6 Disposition of Inhaled Corticosteroids.........................................................................8 Inhalation Device..........................................................................................................9 Physical Characteristics of Patient..............................................................................13 Pharmacokinetic and Pharmacodynamic Properties of ICS.......................................14 Pharmacokinetic Properties.................................................................................14 Prodrug.........................................................................................................14 Mucociliary Clearance.................................................................................15 Pulmonary Residence Time.........................................................................15 Bioavailability..............................................................................................16 Clearance......................................................................................................17 Drug Distribution.........................................................................................18 Half-life........................................................................................................19 Protein Binding............................................................................................19 Objectives of the Study...............................................................................................21 2 IN-VITRO COMPARISION OF FLUTICASONE RESPIRABLE DOSE FROM A METERED DOSE INHALER AND THREE RIGID VALVED HOLDING CHAMBERS..............................................................................................................22 Introduction.................................................................................................................22 Materials and Methods...............................................................................................23 Chemicals and Devices........................................................................................23 Sample Collection...............................................................................................23 Fluticasone Propionate Assay..............................................................................26 v

PAGE 6

Data Analysis.......................................................................................................26 Results.........................................................................................................................27 Discussion and Conclusion.........................................................................................29 3 PLASMA CONCENTRATIONS AND PHARMACOKINETICS PROFILES OF INHALED CORTICOSTEROIDS AND THEIR RELATION TO LUNG FUNCTION IN ASTHMA.........................................................................................32 Introduction.................................................................................................................32 Methods......................................................................................................................33 Subjects................................................................................................................33 Measurements......................................................................................................33 Lung function...............................................................................................33 Plasma drug assays.......................................................................................33 Protocol................................................................................................................34 Analysis...............................................................................................................35 Results.................................................................................................................35 Plasma drug concentrations and pharmacokinetic profiles..........................37 Relation between lung function and AUC...................................................38 Discussion...................................................................................................................40 Plasma Drug Concentrations and Pharmacokinetic Profiles..............................41 Relation Between Lung Function and AUC.......................................................42 Conclusion..................................................................................................................43 4 FLUTICASONE AND BUDESONIDE CONCENTRATIONS AFTER INHALATION EFFECT OF BRONCHOCONSTRICTION....................................45 Introduction.................................................................................................................45 Methods......................................................................................................................46 Subjects................................................................................................................46 Measurements......................................................................................................46 Spirometry....................................................................................................46 Methacholine inhalation challenge...............................................................46 Drug assays..................................................................................................47 Protocol................................................................................................................47 Analysis...............................................................................................................48 Results.........................................................................................................................48 Discussion...................................................................................................................51 5 EVALUATION OF THE ADMINISTRATION TIME EFFECT ON THE CUMULATIVE CORTISOL SUPPRESSION AND CUMULATIVE LYMPHOCYTES SUPPRESSION FOR ONCE-DAILY INHALED CORTICOSTEROIDS................................................................................................55 Introduction.................................................................................................................55 Methods......................................................................................................................57 Historic Data........................................................................................................57 vi

PAGE 7

Population Pharmacokinetic/Pharmacodynamic Analysis..................................58 Pharmacokinetic analysis.............................................................................58 Pharmacokinetic/Pharmacodynamic analysis of effects of BUD and FP on cortisol.................................................................................................59 Pharmacokinetic/Pharmacodynamic analysis of effects of budesonide and fluticasone propionate on lymphocytes.............................................61 Population PK/PD Simulation to Evaluate Administration Time Effect on CCS and CLS for Once-daily Dose of BUD and FP.......................................62 Results.........................................................................................................................63 Population Pharmacokinetic Model....................................................................63 Budesonide...................................................................................................63 Fluticasone propionate.................................................................................65 Population Pharmacokinetic/Pharmacodynamic Model......................................65 Population Pharmacokinetic/Pharmacodynamic Simulation..............................66 Discussion...................................................................................................................73 Conclusion..................................................................................................................78 6 CONCLUSIONS........................................................................................................80 APPENDIX A NONMEM CODE FOR PK MODEL OF BUD AFTER INHALATION.................83 B NONMEM CODE FOR PK MODEL OF FP AFTER INHALATION.....................84 C NONMEM CODE FOR PK/PD MODEL OF BUD..................................................85 D NONMEM CODE FOR PK/PD MODEL OF FP......................................................92 E NONMEM CODE FOR BUD PK/PD SIMULATION MODEL...............................98 F NONMEM CODE FOR FP PK/PD SIMULATION MODEL.................................103 LIST OF REFERENCES.................................................................................................108 BIOGRAPHICAL SKETCH...........................................................................................122 vii

PAGE 8

LIST OF TABLES Table page 3-1 Baseline characteristics............................................................................................36 3-2 Mean (SD) pharmacokinetic parameters for beclometasone monopropionate(BMP),budesonide(BUD), fluticasone propionate(FP) and mometasone furoate (MF)........................................................................................38 3-3 Partial correlation coefficients (r) for the relationship between FEV1 % predicated and AUC after adjusting for age or gender.............................................40 3-4 Predicted AUC values at 50 and 100% predicted FEV1 and PEF and the ratio of these values..............................................................................................................40 4-1 Baseline characteristics............................................................................................49 4-2 Pharmacokinetic parameters for fluticasone and budesonide when inhaled with and without prior methacholine-induced bronchoconstriction.................................50 4-3 Area under the curve (AUC) values for fluticasone and budesonide when inhaled with and without prior methacholine-induced bronchoconstriction............52 5-1 Summary of BUD pharmacokinetic model parameter estimates.............................65 5-2 Summary of FP pharmacokinetic model parameter estimates.................................67 5-3 Summary of FP and BUD pharmacokinetic/pharmacodynamic model parameter estimates...................................................................................................................70 5-4 Summary of the mean cumulative cortisol suppression of the 75 subjects of the 8 simulations...............................................................................................................73 5-5 Summary of the mean cumulative lymphocytes suppression of the 75 subjects of the 8 simulations.......................................................................................................73 viii

PAGE 9

LIST OF FIGURES Figure page 1-1 Molecular structures of several inhaled corticosteroids.............................................2 1-2 Cellular interactions after antigen inhalation in asthma.............................................4 1-3 Classical model of glucocorticoid (GCS) action........................................................8 1-4 Disposition of ICSs after inhalation.........................................................................11 2-1 Schematic diagram of the cascade impactor with aerosol aerodynamic size range at each impactor stage 07 and at the final filter......................................................24 2-2 Mean SD respirable dose of fluticason propionate from metered-dose inhaler (MDI) and valved holding chamber (VHC) devices................................................28 2-3 Mean SD oropharyngeal dose of fluticason propionate from metered-dose inhaler (MDI) and valved holding chamber (VHC) devices....................................29 3-1 Mean (SE) plasma concentration-time curves for beclometasone monopropionate (BMP), budesonide (BUD), fluticasone propionate (FP) and mometasone furoate (MF) following inhalation......................................................37 3-2 Relation between FEV 1 % predicted and the area under the plasma concentration-time curve (AUC) for beclometasone monopropionate (BMP), budesonide (BUD), fluticasone propionate (FP) and mometasone furoate (MF) following inhalation.................................................................................................39 4-1 Mean (SE) plasma drug concentrations following inhalation of 1000 g fluticasone (Accuhaler) and 800 g budesonide (Turbohaler) in 20 subjects with asthma with and without prior methacholine-induced bronchoconstriction....51 5-1 Relationship between population predicted (open triangle) and observed BUD (open circle) concentrations, individual predicted (open square) and observed BUD (open circle) concentrations as function of time.............................................64 5-2 Relationship between population predicted (open triangle) and observed FP (open circle) concentrations, individual predicted (open square) and observed FP (open circle) concentrations as function of time......................................................66 ix

PAGE 10

5-3 Diagnostic plots of BUD PK/PD model...................................................................68 5-4 Diagnostic plots of FP PK/PD model.......................................................................69 5-5 Relationship between administration time and corresponding mean (+ SE) cumulative suppression............................................................................................71 5-6 Relationship between administration time and corresponding mean (+ SE) cumulative suppression............................................................................................72 x

PAGE 11

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FACTORS AFFECTING PULMONARY TARGETING OF INHALED CORTICOSTEROIDS By Kai Wu December 2006 Chair: Guenther Hocchaus Major: Pharmaceutical Sciences Understanding the factor affecting pulmonary targeting would help us improve inhaled corticosteroids (ICSs) therapy. The three factors important for pulmonary selectivity are inhalation device, characteristics of patients, and pharmacokinetic/pharmacodynamic properties of ICSs. Certain aspects related to these factors were evaluated in the current dissertation. The effect of different valved holding chambers on in vitro aerosol deposition from a fluticasone MDI was assessed using an in vitro cascade impactor method. It was found that the mean fluticasone respirable dose from AeroChamber-Plus and OptiChamber was no different (p > 0.05) from the respirable dose produced by the MDI alone, while the OptiChamber -Advantage respirable dose was significantly less compared with the dose from the MDI alone (p < 0.05). The overall goal of the second and third study was to test the hypothesis that the pulmonary systemic availability of ICSs is related to lung function of the patients while xi

PAGE 12

the extent of lung function affecting the bioavailability depends on the physicochemical properties of ICSs. In the second study, it was found that the bioavailability for all four drugs (beclomethasone, budesonide, mometasone and fluticasone) tended to be higher in the asthmatic patients with higher lung function, while the associations between the bioavailability and lung function differed for the four drugs. In the third study, it was found that AUC 0-5hour values for fluticasone and budesonide were lower by a median of 60% and 29% respectively when bronchoconstriction was induced in the asthmatic patients. Finally, the effect of administration time on cumulative cortisol (CCS) and cumulative lymphocytes suppression (CLS) was evaluated for once-daily ICS and a population pharmacokinetic/pharmacodynamic modeling/simulation approach was used. It was found that the optimal time for administration of ICS depends on the choice of the biomarker: afternoon in terms of minimized CCS, and morning in terms of minimized CLS. xii

PAGE 13

CHAPTER 1 INTRODUCTION More than 7% of adults and 12% of children in the United States (US) are affected by asthma, a chronic inflammatory disease of the airways [1]. Over the world, asthma afflicts 100 to 150 million people. The prevalence of asthma in the US has been increasing by 5% to 6% over the past decades [2]. Asthma is responsible for 100 million days of restricted activity and 470,000 hospitalizations each year [1]. In 1998, asthma related cost was estimated at $11.3 billion in the U.S. alone. Inhaled corticosteroids (ICSs) have revolutionized the treatment of asthma in the past 30 years and now represent the first-line therapy for patients with chronic disease [3]. Numerous clinical studies have demonstrated the high anti-asthmatic efficacy (e.g. prevention and control of symptoms) and reduced systemic side effects of ICS therapy. Introduction of beclometasone dipropionate in 1972 as a pressurized metered-dose inhaler (MDI) marked the beginning of the targeting treatment for asthma [4]. Since then, other potent topical corticosteroids have been developed, and currently flunisolide (FLU), triamcinolone acetonide (TA), budesonide (BUD), fluticasone propionate (FP), mometasone furoate along with BDP (Figure 1-1) are available in the US. Additionally, ciclesonide (Figure 1-1) is another ICS currently under development for the treatment of asthma of all severities. The use of inhaled corticosteroids for asthma allows, in effect, local delivery of the drug to the lung. The goal of inhaled corticosteroids is to produce long-lasting therapeutic effects at the pulmonary target site and to minimize systemic effects (pulmonary 1

PAGE 14

2 selectivity). In general, pulmonary targeting of ICS is determined by their pharmacokinetic and pharmacodynamic (PK/PD) properties, the biopharmaceutical aspects of the delivery device and the characteristics of the patient. In this chapter, an overview of factors affecting pulmonary targeting of ICS will be discussed. In addition, the current literature on asthma pathogenesis and the mechanism of action of ICS in asthma therapy is reviewed. Figure 1-1. Molecular structures of several inhaled corticosteroids [5, 6].

PAGE 15

3 Asthma Asthma is a chronic disorder of the airways characterized by airway inflammation, reversible airflow obstruction and airway hyperresponsiveness. Studies have shown that airway inflammation is an integral element in the pathogenesis of asthma and a driving force in airway hyperresponsiveness and a contributing factor for variable levels of airflow obstruction [7]. Asthma inflammation involves many cells, with mast cells, eosinophils, basophils, neutrophils, macrophages, epithelial cells, and lymphocytes all being active participants (Figure 1-2). The interactions of these cells result in the release of different mediators leading to bronchoconstriction, microvascular leakage, mucus hypersecretion, epithelial damage, and stimulation of neural reflex. For example, activation of the mast cell by specific antigen through cell-bound IgE releases histamine and causes synthesis of leukotrienes, which can precipitate an acute episode of airway obstruction [8]. Mast cells also release inflammatory enzymes or proteases, which can contribute to airway inflammation and airway hyperresponsiveness. Finally, mast cells express and release a wide variety of proinflammatory cytokines including granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin (IL)-3, 4 and 5 [9]. These cytokines change the immunologic environment of the airway and are likely instrumental in initiating the inflammatory response in asthma through changes in adhesion proteins, cell recruitment, and inflammatory cell survival. Eosinophils are recruited to the lung in asthma after antigen inhalation and are the main effector cells in the resulting inflammatory process. Eosinophils contribute to inflammation by release of mediators (i.e., leukotrienes and toxic oxygen products), by degranulation and release of granular proteins, and by generation of cytokines [8, 10]. Granular proteins, such as major basic protein, directly

PAGE 16

4 damage airway tissue, promote airway responsiveness, and injure airway epithelium. In general, the degree of airway eosinophilia is usually proportional to the severity of asthma [11]. Activated lymphocytes are another important element of airway inflammation. One of the T helper (T H ) lymphocytes, T H2 shows a cytokine profile that contains IL-4 and IL-5, two cytokines that are essential for IgE synthesis (IL-4) and eosinophil production (IL-5). Since lymphocytes are long lived, they are believed to be crucial to the persistence of the inflammatory process in asthma. Epithelial cells are another source of proinflammatory mediators in asthma including eicosanoids, cytokines, chemokines, and nitric oxide [12]. Figure 1-2. Cellular interactions after antigen inhalation in asthma [13] The current concept of asthma therapy is based on a stepwise approach, depending on disease severity, and the aim is to reduce the symptoms that result from airway obstruction and inflammation, to prevent exacerbations and to maintain normal lung function [14]. Controlling the allergic response to the local environment due to smoking and allergens found at home or in the workplace, is the first priority in minimizing

PAGE 17

5 asthma hyperresponsiveness. Traditionally, there are five classes of medications for asthma therapy: corticosteroids (e.g., beclomethasone dipropionate), short-acting (e.g., albuterol) and long-acting (e.g., salmeterol) beta-agonists, theophylline, cromolyn sodium and nedocromil sodium, and anticholinergics (e.g., ipratropium bromide) [15]. Corticosteroids, cromolyn sodium and nedocromil sodium can be classified as anti-inflammatory medications. Beta-agonists and theophylline are both brochodilators. Anticholinergics work by blocking the contraction of the underlying smooth muscle of the bronchi. Among these medications, inhaled corticosteroids are the most effective for the treatment of asthma. Their efficacy is related to many factors including a diminution in inflammatory cell function and activation, stabilization of vascular leakage, a decrease in mucus production, and an increase in beta-adrenergic response. The strong scientific rationale behind the combination therapy of long-acting beta-aganists and corticosteroids has led to the development of two mixed combination inhalers: salmeterol/fluticasone and formoterol/budesonide. They are increasingly used as a convenient controller in patients with persistent asthma. Advances in the knowledge of the pathophysiology of asthma in the past decade have opened the way to the development of novel therapeutic strategies that target more directly on the pathophysiology of asthma than those currently in use [16]. Various compounds able to interfere with the complex network of proinflammatory mediators, cytokines, chemokines, and adhesion molecules involved in the pathogenesis of asthma have been identified, such as leukotriene modifiers, tachykinin antagonists, endothelin antagonists, adenosine receptor inhibitors, tryptase inhibitors, cytokine and chemokine inhibitors, direct inhibitors of T-cell function and adhesion molecule blockers. New forms of immunotherapy, such as anti-IgE therapies,

PAGE 18

6 gene vaccination with plasmid DNA and mycobacterial preparations, are also emerging to aim at blocking the unbalanced T H2 response that characterizes the pathophysiology of asthma. Despite the significant expansion of the experimentally available treatment options, only the leukotriene modifiers are on the market, among which leukotriene antagonists are now most frequently used in asthma treatment [17, 18]. The leukotriene antagonists have been shown to inhibit exercise-provoked bronchospasm, to modify the airway reponse to inhaled antigen, and to improve airway function in patients with chronic asthma. But in head-to-head trials with inhaled corticosteroids, the leukotriene antagonists are less effective in terms of improvement in lung function and reduction in exacerbations [7]. Thus far, inhaled corticosteroids still represent the cornerstone of treatment for asthma. Mechanism of Action of Corticosteroids in Asthma Corticosteroids are able to affect many of the inflammatory pathways involved in the pathogenesis of asthma, including the complex cell to cell communications mediated by the so-called "cytokine network" [19]. Corticosteroids start their action when the lipophilic corticosteroid molecule crosses the cell membrane and binds to a specific, intracellular glucocorticoid receptor (GR) (Figure 1-3). The GR is located in the cytoplasma and belongs to the steroid thyroid/retinoic acid receptor superfamily [20]. Prior to binding corticosteroids, GR forms a large heteromeric complex with several other proteins, from which it dissociates upon ligand binding. 90 kDa heat shock protein (hsp90) plays a central role in this complex [21]. It binds as a dimer to the C-terminal ligand binding domain (Figure 1-3). The association with hsp90 has been shown to be required to maintain the C-terminal domain of GR in a favorable conformation for ligand binding [22], but GR is transcriptionally inactive at this stage. Binding of corticosteroids

PAGE 19

7 to GR induces dissociating hsp90 and other heat shock proteins, resulting in a conformational change which allows the GR complex to translocate to the nucleus or interact with cytoplasmic transcription factors. Once in the nucleus, the GR complex binds as a dimer to specific DNA sites, a specific nucleotide palindromic sequences termed glucocorticoid response elements (GRE) (Figure 1-3) [23]. Then the transcription of specific genes can be increased (transactivation) or decreased (transrepression) depending on whether the GRE is positive or negative. In particular, corticosteroids increase the expression of anti-inflammatory proteins such as lipocortin-1, interleukin-1 receptor antagonist (IL-1ra), interleukin-10 (IL-10), secretory leukocyte inhibitory protein, neutral endopetidase and the inhibitory protein (IB) of nuclear factor-B (NF-B) [3]. It is known that lipocortin-1 inactivates the enzyme phospholipase A 2 thus inhibiting the production of lipid mediators (i.e., platelet-activating factor, leukotrienes, and prostaglandins) and NF-B positively regulated inflammatory and immune responses [24, 25]. Corticosteroids are also able to enhance the transcription of the gene encoding the 2 -adrenergic receptor, thus reversing and/or preventing the down-regulation possibly induced by long-term treatments with 2 -agonist bronchodilators [13]. However, the genomic mechanism of transactivation is believed to be mainly responsible for the unwanted side effects of corticosteroids, rather than for their anti-inflammatory and immunoregulatory actions. The very effective control of airway inflammation exerted by corticosteroids in asthma is largely mediated by inhibition of the transcriptional activity of several different genes encoding pro-inflammatory proteins such as cytokines (IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-11, IL-13, TNF, GM-CSF), chemokines (IL-8, RANTES, MIP-1 MCP-1, MCP-3, MCP-4, eotaxin), adhesion

PAGE 20

8 molecules (ICAM-1, VCAM-1, E-selectin), and mediator-synthesizing enzymes (i-NOS, COX-2, cytoplasmic PLA 2 ), which are mostly due to GR-dependent repression of pro-inflammatory genes [19]. Corticosteroids also exert their repressive action by direct interaction with transcription factors and post-transcriptional regulation [19]. Figure 1-3. Classical model of glucocorticoid (GCS) action. The glucocorticoid enters the cell and binds to a cytoplasmic glucocorticoid receptor (GR) that is complexed with two molecules of a 90kDa heat shock protein (hsp 90). GR translocates to the nucleus where, as a dimer, it binds to a glucocorticoid recognition sequence (GRE) on the 5-upstream promoter sequence of steroid-responsive genes. GREs may increase transcription and nGREs may decrease transcription, resulting in increased or decreased messenger RNA (mRNA) and protein synthesis (Adapted from Barnes [26]). Disposition of Inhaled Corticosteroids Understanding the disposition of ICS after inhaled administration will help us understand why PK/PD properties, the choice of delivery device and the physical characteristics of the patient are important for achieving pulmonary targeting of ICS (Figure 1-4).

PAGE 21

9 Upon inhalation of a glucocorticoid, only 10% to 60% of the administered dose is deposited in the lungs. The other of the dose (40% to 90%) impacts on the oropharyngeal region and is swallowed. The swallowed dose is absorbed from the gastrointestinal tract and after undergoing first pass inactivation by the liver only orally bioavailable fraction enters the systemic circulation. The drug particles deposited in the lungs dissolve according to the rate of dissolution of the drug. If not readily dissolved, the drug will be removed by mucociliary clearance back into the oropharyngeal region and then is swallowed. Only the dissolved drug interacts with the glucocorticoid receptors (GR), as described in the previous section, present in the lung to produce the anti-inflammatory effects. Following the interaction with the GR, the drug is absorbed into the systemic circulation. The absorption processes are rapid and it is believed that dissolution is the rate-limiting step involved in pulmonary absorption of IGCs. 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 GR to produce the systemic side effects. The drug is eventually eliminated from the systemic circulation mainly by hepatic clearance mechanisms. Inhalation Device One factor to be considered for distinct pulmonary selectivity is efficient delivery to the lungs. As mentioned earlier, about 10% to 60% of nominal inhaled dose will deposit in the lung. The exact percentage and the distribution of the dose in the lung depends great deal on the biopharmaceutical aspects of the delivery device. It also depends on the patient since every inhaler requires certain level of technique for optimal delivery.

PAGE 22

10 There are three main classes of inhalation devices for ICSs: metered-dose inhaler (MDI), breath-activated dry powder inhaler (DPI) and nebulizer. Inside MDI the drug exists as a suspension in a carrier liquid or a solution delivered through a chlorofluorocarbon (CFC) or hydrofluoroalkane (HFA) propellant, respectively [27]. However, CFC-MDIs are gradually phased out because of their ozone-depleting potential. In addition to their environmentally friendly property, HFA solutions also seem to have the advantage of delivering a much greater mass of fine particles, with a diameter of less than 5 m. Fine particles are more likely to be deposited in the tracheo-bronchial and pulmonary regions in the lung. On the other side, larger particles are deposited mostly in the oropharynx where they are swallowed and increase the risk of systemic absorption [28]. The average particle diameter delivered by a CFC-MDI is 3.5-4.0 m whereas the average particle diameter delivered by a HFA propellant is around 1.1 m. This difference in particle diameter might have a clinical significance as the average diameter of small airways is around 2m, resulting in a greater lung deposition [29]. This increased proportion of fine particles with the HFA-MDI results in an improved lung deposition. A previous study showed the lung deposition of beclometasone dipropionate (BDP) was increased from 4-7% using CFC-MDI to 55-60% using a newly developed HFA-MDI [30]. Another study showed with the CFC-free solution MDI, a mean lung deposition of 52% for ciclesonide could be obtained [31]. In a single-dose study comparing HFA flunisolide and CFC flunisolide, the drug deposition in the lung could even be increased to 68% (HFA) compared to 19.7% (CFC) [32]. Hence, replacing the propellant CFC with HFA leads to an increase in lung deposition. Lung deposition can also be increased by use of spacer and valved holding chamber(VHC) devices, which can

PAGE 23

11 alter the amount of fine particles and therefore, increase the respirable fraction and decrease the amount of drug deposited in the oropharynx, [33]. Moreover, use of VHC with MDI helps the patients who lack coordination between actuation and inhalation. However, it also needs to be kept in mind that a greater lung deposition might result in a greater possibility of systemic adverse effects because of the lack of first-pass metabolism after direct absorption from the lung. Mouth and Pharynx Mucociliary clearanceLung 10 -60 % LungDeposition GI tract 40 -90 % swallowed Absorptionfrom Gut Liver First-passInactivation Pulmonary Absorption SystemicCirculationOrally absorbedFraction Clearance SystemicSide Effects Clinical effectsMouth and Pharynx Mucociliary clearanceLung 10 -60 % LungDeposition GI tract 40 -90 % swallowed Absorptionfrom Gut Liver First-passInactivation Pulmonary Absorption SystemicCirculationOrally absorbedFraction Clearance SystemicSide Effects Clinical effects Figure 1-4. Disposition of ICSs after inhalation (Adapted from [34]) The other inhaler type used for ICSs is the dry powder inhaler. Use of DPI requires easier delivering technique that requires less coordination than the MDI. However, it requires a forceful deep inhalation to trigger the inhalation device to help break up the aggregates of the micronized powders into respirable particles in the oropharynx and larger airways. Therefore, lung deposition is flow-dependent and the higher the inhalation flow, the smaller the particles will be [35]. An inspiratory flow of 60L/min is considered

PAGE 24

12 to be optimal [33]. Hence, it should be ensured that asthmatic patients in all asthma stages are able to achieve an inhalation flow that is enough to achieve the required effect [35]. In a lung deposition study it was shown that reducing the inhalation flow from 58 L/min to 36 L/min reduces the lung deposition of budesonide from around 28% to around 15% [36]. The standard and most common used jet nebulizer is based on a constant output design, run by compressed air or oxygen with supplemental air drawn in across the top of the nebulizer. The nebulizer contains the drug either as a suspension or a solution. There have been also new developments in the field of nebulizers and liquid formulations. Among those are the inhalation device Mystic from Batelle, which is based on electro-hydrodynamic principles, employing electrostatic energy to create fine aerosols from formulated drug solutions or suspensions thereby increasing the pulmonary tract deposition to about 80% [37] and the RESPIMAT device from Boehringer Ingelheim uses a high-pressure micro-spray system of nozzles to release a metered dose to the patient. This system generates a slow release of the drug with a high concentration of respirable particles[38]. Improved pulmonary deposition will lead to improved pulmonary targeting, especially for drugs with significant oral bioavailability such as BDP, as efficient pulmonary delivery will reduce the amount of swallowed drug and, subsequently, oral absorption. In addition, the new more efficient pulmonary delivery devices reduce oral deposition which may also reduce oropharyngeal adverse effects.[39] However, efficient pulmonary deposition, with respect to systemic adverse events, is not as important for

PAGE 25

13 drugs with low oral bioavailability, such as ciclesonide, mometasone furoate and fluticasone propionate, as very little drug will be absorbed through the oral route anyway. Physical Characteristics of Patient Many aspects of physical characteristics of the patient could affect pulmonary targeting of a ICS therapy. One of them is age. Usually, a proper technique for use of inhalation device is much easier to be learned and employed among adults. On the contrary, children and elderly might not learn to use the device properly, which could leads to less or more dose than right dose taken and reduced pulmonary selectivity of ICS therapy. The other issue related with age is the compliance to dosing regimen. In a study evaluating compliance with ICSs, it was found that the children between 8 and 12 years only averaged 58% of compliance [40]. Furthermore, only 32% of those doses were actually received at the correct time. A non-compliance with dose is highly likely to associated less efficacy of the therapy. Moreover, non-compliance with timing of the dose would probably affect the pulmonary selectivity of therapy, as it has been shown by some other studies that administration time is an important factor to maximize the efficacy and minimize the systemic side effects of ICSs [41-48] The important of administration time on systemic side effects of ICSs will be discussed in later chapter of this dissertation. Another important aspect of patient characteristics affecting pulmonary selectivity of ICSs is the lung function. It has been suggested that the pulmonary systemic bioavailability of certain ICSs is associated with the lung function of patient [49]. In one study with a single dose fluticasone dry powder in patients with asthma of varying degrees of severity there was a highly significant linear correlation between the absolute magnitude of adrenal suppression and the lung function expressed as percentage

PAGE 26

14 predicted forced expiratory volume in 1 second (FEV 1 ) [50]. This is consistent with pharmacokinetic data where there was 62% lower plasma fluticasone concentrations (as AUC) in patients with moderately severe asthma (FEV 1 = 54% predicted) than in healthy volunteers receiving inhaled fluticasone. However, in another study, there was no difference in plasma budesonide concentrations between healthy volunteers and patients with mild asthma after inhalation budesonide, a much less lipophilic compound than fluticasone [51]. The correlation between lung bioavailability, lung function and drugs lipophilicity will be discussed in later chapters. Pharmacokinetic and Pharmacodynamic Properties of ICS Pharmacokinetic Properties Prodrug A prodrug is a pharmacologically inactive compound that is activated in the body after its administration. To exert a local effect, a prodrug needs to be activated in the target tissue, e.g. lung. As an example, the ester functional group in ciclesonide needs to be cleaved by esterases in the lungs, before the active compound desisobutyryl-ciclesonide can interact with the receptor [6, 52]. The use of a prodrug might be beneficial as high lipophilicity may increase the pulmonary residence time. However, activation predominantly in the lungs is required to maintain pulmonary selectivity. One additional advantage of drugs that are delivered in an inactive form is that they might reduce certain oropharyngeal adverse effects such as oral candidiasis. It has been shown that the negligible activation of ciclesonide in the oropharyngeal region reduce the oropharyngeal adverse effects [39, 53].

PAGE 27

15 Mucociliary Clearance The mucociliary transporter is able to remove those slowly dissolving solid drug particles deposited to the upper part of the respiration tract. This phenomenon will be less pronounced for rapidly dissolving ICS or those present in solution. Mucociliary clearance mechanism is more pronounced in the central airways of the lung compared to the peripheral airways and plays an important role in reducing the overall availability (systemic + pulmonary) of highly lipophilic glucocorticoids that have a slow rate of dissolution. Studies have shown that the systemic availability of more slowly dissolving ICS, such as fluticasone propionate, is lower in patients with asthma when compared with healthy volunteers [54, 55]. The reason is that deposition in patients with asthma tends to be more central in the lungs. Pulmonary Residence Time The large surface area, thin membranes and the existence of pores allow an efficient exchange between the outer part of the lung and the circulatory system. Once dissolved in the lung lining fluid, ICS, as lipophilic drugs, will easily cross cell membranes and be absorbed quickly and efficiently into the systemic circulation [56]. Therefore, to achieve pulmonary selectivity requires the drug efficiently delivered to the lung, however a fast absorption into the systemic circulation immediately after delivery shall be prevented. Triamcinolone acetonide, when given as solution into rat lungs, was unable to produce pulmonary targeting as the receptor occupancy in lung and liver was very similar [57]. Extending the residence time of the drug in the lung (e.g. through controlling the release of the drug in the lung) will increase pulmonary selectivity as high drug concentrations are present in the lung for a longer period of time [58]. Various approaches to extend the pulmonary residence time of drugs have been evaluated in the

PAGE 28

16 pre-clinical setting.[59, 60] To achieve increased pulmonary residence time, two approaches are currently utilized for ICSs. Highly lipophilic drugs, such as fluticasone propionate, will dissolve slowly after deposition in the lung as particles and insure high lung concentrations over a longer period of time. A pharmacokinetic parameter used to quantify pulmonary residence time is the mean pulmonary absorption time (the average time the drug will need to leave the lung). Commercially available drugs differ in this property, with fluticasone propionate showing the highest pulmonary residence time of almost 5 hours while flunisolide is absorbed very rapidly. An alternative biological slow release system has been identified for budesonide and potentially ciclesonide [61-63]. In pulmonary cells, these two drugs form lipophilic esters that are biologically inactive but are so lipophilic that they are unable to cross membranes therefore they are trapped inside the cell to form a reservoir of active drug since the esters will be slowly cleaved back to the active drug. This esterfication/ester-cleavage system, therefore, ensures a long residence time in the lung which might provide sufficiently prolonged asthma control for a reduced dosing regimen with budesonide and ciclesonide. Bioavailability The bioavailability of an inhaled corticosteroid is the rate and extent at which the drug reaches its site of action (pulmonary bioavailability) as well as the blood (systemic bioavailability). As mentioned earlier, a large fraction (approximately 40-90%) of the inhaled dose is swallowed and subsequently available for systemic absorption. This bioavailability of the orally delivered part is dependent on absorption characteristics of the drug from the gastro-intestinal tract and the extent of intestinal and hepatic first-pass metabolism. The oral bioavailabilities of currently used corticosteroids range from less than 1% for fluticasone propionate to 26% for 17-beclomethasone monopropionate [64

PAGE 29

17 70]. However, the main determinant of systemic bioavailability after inhalation is direct absorption from the lung, where for the currently available inhaled corticosteroids there is no first-pass effect. All of the drug that is deposited in the lung will be absorbed systemically [49]. The percentage of the drug dose that is deposited in the lung is greatly influenced by the efficiency of the delivering device. Since the orally absorbed fraction of the drug does not contribute to the beneficial effects but can induce systemic side effects, it is desirable for the oral bioavailability of inhaled corticosteroids to be very low. Therefore, the pulmonary bioavailability is rather a function of the delivery device used for inhalation than a property of the drug itself. The pulmonary bioavailability will depend on the amount deposited in the lungs and will differ with the delivery device used [49, 71]. Fluticasone propionate, for example, has an oral bioavailability of <1% due to a high first-pass metabolism (see above). When administered to the lungs using a dry powder inhaler (DPI), the absolute bioavailability (systemic + pulmonary) is reported to be approximately 17% and compared to 26% to 29% when using a metered dose inhaler (MDI) [27, 72]. After mometasone fuorate administration via a dry powder inhaler the absolute bioavailability was reported to be 11% [72]. Clearance Clearance is a PK parameter describing the efficiency of the body to eliminate the drug and quantifies the volume of the body fluid that is cleared of drug over a given time period. For most drugs showing linear protein binding and non-saturated elimination this value is a constant. The higher the clearance of an ICS the lower the systemic adverse effects will be and the more pronounced the pulmonary selectivity. If drugs are mainly metabolized and cleared by the liver, the maximum clearance possible is that of the liver blood flow of ~90 L/hour. Several of the available ICS (budesonide, fluticasone

PAGE 30

18 propionate) show clearance values close to the liver blood flow [73, 74]. The clearance of such high extraction drugs is independent of protein binding. The active metabolites of ciclesonide and beclomethasone propionate have been reported to express higher clearance values when given intravenously [65, 75]. While these data suggests extra-hepatic metabolism, future studies with the active metabolites need to be performed to clearly confirm the high clearance of these two drugs, especially as this would represent a clear advantage for these drugs over other commercially ICS. Drug Distribution The volume of distribution (V d ) is a PK parameter that describes the distribution of drug in the body. The higher the V d the greater the concentration of drug present in the tissues and the lower the plasma concentration. For drugs such as ICS that can cross membranes easily, the V d is determined primarily by how much drug is bound to plasma proteins and how much drug is bound to tissue components. Thus, corticosteroids with a very large volume of distribution (300-900L) are extensively distributed and bound to the tissues. The more drug that is bound in the tissue compartment (the smaller the fraction unbound in tissue f ut ,) the higher will be the V d Similarly, a weak plasma protein binding (fraction unbound drug in plasma f u is large) will allow more drug to enter the tissue compartment. Because the V d is determined by the ratio of protein binding in plasma and tissue, it is possible for two drugs with significant differences in the plasma protein binding to show the same V d This indicates that the V d is a parameter that is difficult to evaluate. Computer simulations did show that an increase in V d although affecting the half-life of a drug, does not affect the pulmonary selectivity [76].

PAGE 31

19 Half-life Half-life (t 1/2 ) is a secondary PK parameter that is determined by clearance and the V d [77]: t 1/2 = 0.693 V d /clearance A short half-life, because of a high clearance, is beneficial for achieving pronounced pulmonary selectivity. However, a short half-life due to a small V d does not affect the pulmonary selectivity [76]. The half-life is therefore of little value if an ICS has to be evaluated with respect to its pulmonary selectivity. Protein Binding Protein binding is one of the important PK properties of ICS. The major protein that is involved in binding with ICSs is albumin. Albumin comprises about 55-62% of the protein present in plasma and also occurs plentifully in the tissues of mammals. Protein binding is seen as storage form of the drug because it is reversible and a rapid equilibrium is established. High plasma protein binding is considered as a desirable characteristic for ICSs. Once ICSs reach the systemic circulation, binding to plasma proteins decreases the free unbound fraction in the systemic circulation and thus decreases the risk for systemic side effects, since only the free, unbound drug is pharmacologically active. Reported clinical studies showed that ciclesonide, a novel inhaled corticosteroid with plasma protein binding of 99% for itself and its active metabolite had an incidence of adverse effects that was not significantly greater than that observed with placebo [31, 78, 79]. The extent of plasma protein binding differs among ICSs and it has been positively associated with their lipophilicity. Current ICSs in development tend to be relatively lipophilic. Plasma protein binding has been reported to be 80% for flunisolide, 87% for beclomethasone dipropionate, 88% for budesonide, 71% for triamcinolone acetonide,

PAGE 32

20 90% for fluticasone propionate, 98% for mometasone furoate and 99% for active principle of ciclesonide [79-82]. As discussed above, those ICSs with higher plasma protein binding would have a safety margin over those ICSs with lower plasma protein binding. However, the increased plasma protein binding may not necessarily improve the risk-benefit ratio. As lipophilic compound, ICSs could penetrate cells and bind to both plasma and tissue proteins. It is reasonable to expect the tissue protein binding for those ICSs with high plasma protein binding would be high too because of their high lipophilicity. The fraction of free drug at the target site would be reduced. The high tissue protein binding is also indicated by ICSs high volume of distribution. The steady state volume of distribution (Vss) of a drug can be calculated according to: tutuppssVffVV where f up is the unbound fraction of drug in plasma, f ut is the unbound fraction of drug in tissue, Vp is the plasma volume and Vt is the volume of tissue compartment. The Vss of ICSs could also be positively associated with their lipophilicity. Thus, for those relative lipophilic IG with higher plasma protein binding and Vss, as we can see from above equation, they would have increased tissue protein binding as well. An increase in tissue protein binding of just a few percent can translate into a significant reduction in the amount of efficacy. For example, if one was 99% bound in the lung and another was about 98% bound in the lung, the one with 98% bound would present two folds of unbound concentration for interaction with receptor to produce beneficial effects. A higher dose may be needed for the ICSs with higher tissue binding to achieve efficacious free concentration, thus the safety margin obtained by their high plasma protein binding

PAGE 33

21 would be offset. Therefore, it is essential to determine both plasma protein binding and tissue binding, especially the target tissue, lung protein binding for a complete evaluation of efficacy and safety profile of ICSs. Objectives of the Study Pulmonary targeting is essential for successful inhaled corticosteroids (ICS). Many factors determine the pulmonary targeting of ICS and these factors include pharmacokinetic and pharmacodynamic (PK/PD) properties of ICS, inhalation device and patients characteristics. Understanding of these factors has contributed significantly to improve pulmonary targeting of modern ICS. This research has been focusing on evaluation of some aspects related to these factors and the overall objectives could be summarized as followings: Aim 1: Compare the in-vitro respirable doses of fluticasone from a metered dose inhaler and three rigid valved holding chambers. Aim 2: Assess pharmacokinetic profiles of inhaled beclomethasone, budesonide, fluticasone and mometasone in asthmatic subjects with a broad range of lung function. Aim 3: Compare the systemic exposure change between inhaled budesonide and fluticasone after lung function decrease induced by methacholine in asthmatic subjects. Aim 4: Evaluate administration time effect on the cortisol suppression and lymphocytes suppression for once-daily inhaled corticosteroids.

PAGE 34

CHAPTER 2 IN-VITRO COMPARISION OF FLUTICASONE RESPIRABLE DOSE FROM A METERED DOSE INHALER AND THREE RIGID VALVED HOLDING CHAMBERS Introduction There are several metered-dose inhaler (MDI) valved holding chambers (VHCs) available in the United States. The primary purpose of VHC is to hold the MDI puff very shortly before inhalation to minimize the detrimental effects of poor timing between MDI actuation and patient inhalation. Another important feature of VHC is to allow the MDI propellant time and distance to evaporate after actuation. Evaporation of propellant promotes the formation of smaller aerosol particles that are likely to be carried into small airways. Moreover, VHC collect large aerosol particles that would otherwise deposit into the oropharynx. VHC has been recommended for all patients who can not effectively use MDI alone, such as children and for all asthma patients prescribed an ICS to reduce the risk of topical side effects of corticosteroid such as thrush and dysphonia [1]. In vitro aerosol data indicated that the VHC choice has no impact on the aerosol characteristics of bronchodilators from an MDI [83]. Further, clinical studies have been unable to discern measurable difference in bronchodilator effect among various albuterol MDI-VHC combinations [84, 85]. Hence, many clinicians and third-party payers conclude that all VHCs available in the USA markets are equivalent. However, this statement may not stand for ICS as several in vitro studies indicate that the quantity of ICS available for delivery to human lung can vary as many as three folds, depending on the choice of VHC [83]. Therefore, interchanging one VHC with another may result in 22

PAGE 35

23 pronounced fluctuations in the aerosol dose reaching the lungs and potentially result in therapeutic failures. Fluticasone propionate MDI (Flovent, GlaxoSmithKline, Research Triangle Park, NC) is the most often prescribed ICS in the USA for asthmatic patients. Hence, to assess the aerosol characteristics of fluticasone propionate from various VHCs is very important for delivery of successful therapy to asthmatic patients. Materials and Methods Chemicals and Devices Acentone nitrile and trifluoroacetic acid were obtained from Fisher Scientific. Beclomethasone dipropionate was kindly provided by 3M. Double distilled deionized water was prepared in our lab (Gainesville, FL). The MDI tested was 110g Flovent (GlaxoSmithKline, Research Triangle Park, NC). The VHCs tested were AeroChamber-Plus (Monaghan Medical, Plattsburgh, NY), OptiChamber (Respironics, Murrysville, PA) and OptiChamber -Advantage (Respironics, Murrysville, PA). Andersen Mark-, eight-stage cascade impactor equipped with a USP induction port was purchased from Thermo Andersen (Smyrna, GA). Sample Collection The deposition characteristics of fluticasone propionate aerosol delivered from the MDI alone (110g Flovent ), the MDI attached to AeroChamber-Plus VHC, OptiChamber VHC, and OptiChamber -Advantage VHC were examined by using the well-established in-vitro cascade impactor method [86, 87]. Fluticasone propionate aerosol particles that entered the cascade impactor were sorted by aerodynamic diameter onto eight different stages (Figure 2-1).

PAGE 36

24 Figure 2-1. Schematic diagram of the cascade impactor with aerosol aerodynamic size range at each impactor stage 0 and at the final filter (F). MDI = metered-dose inhaler; VHC = valved holding chamber; USP = United States Pharmacopeia. Aerodynamic diameter does not refer to actual aerosol droplet diameter, rather it is a means of categorizing particles that have different shapes and densities with a single dimension based on how those particles behave when settling out. Aerosol deposition onto the throat intake port was reflective of in vivo oropharyngeal deposition [88]. Aerosol deposition on stages 0, 1, and 2 of the cascade impactor corresponded to particles of 5.1 m. These particles are too large for therapeutic use in asthma and likely would deposit into the large upper airways such as the trachea [89-91]. Aerosol that deposited on stages 3, 4, and 5 of the cascade impactor corresponded to particles of 1 m, a range considered ideal for deposition and retention within the small human airways (bronchi and bronchioles) [89]. Aerosol collected on stages 6, 7, and in the final filter of the cascade impactor corresponded to particles smaller than 1 m, a range considered too

PAGE 37

25 small for therapeutic use in asthma and one that favors deposition into the respiratory bronchioles and alveoli [89]. Fluticasone propionate aerosol from the MDI connected to each of six spacer configurations and aerosol from the MDI alone were sampled directly into the cascade impactor (Mark IIACFM eight-stage nonviable ambient sampler; Andersen Instruments Inc., Atlanta, GA) through an artificial throat manufactured according to United States Pharmacopeia (USP) standards [87]. Aerosol was drawn into the impactor by continuous airflow generated by a vacuum pump (model 20-709; Graseby-Andersen, Atlanta, GA). Airflow through the impactor was calibrated to conform to the manufacturers requirements (28.3 L/min 5%) and simulated a tidal inspiratory flow of about 30 L/minute. Before each sampling run, continuous airflow through the impactor was allowed to equilibrate for 5 minutes, and each fluticasone propionate MDI canister was primed 5 times in a separate room to ensure uniformity among actuations. Before each sampling run, the MDI valve was washed twice with methanol and dried to remove any trace drug that resulted from priming. Each MDI canister was shaken for at least 10 seconds before each actuation. Each of the above devices was tested five times, and during each run, 5 MDI puffs were sampled through the cascade impactor without delay between actuation and sampling. After each actuation into the impactor, airflow was maintained for 60 seconds before the next actuation. After the final actuation in each run, flow through the impactor was maintained for an additional 60 seconds. Laboratory personnel wore antistatic outer clothing containing carbon fibers during sample collection to minimize transfer of electrostatic charges to the VHCs. All VHCs were washed in a liquid detergent solution and allowed to air dry before use, to minimize electrostatic

PAGE 38

26 attraction of fluticasone propionate aerosol during sample collection [92]. Ambient temperature (22 2C) and humidity (46 4%) in the aerosol laboratory were monitored to minimize inter-day variability in sample collection Fluticasone Propionate Assay After each trial, fluticasone propionate was washed from the USP throat and from each of eight collection plates in the cascade impactor with methanol and an equal volume of distilled water. The volume used to wash each section of the apparatus varied from 5 ml to eliminate concentrations of fluticasone propionate below the assay limit of detection (0.5 g/ml). Two 1-ml aliquots of the resultant solutions were collected for high-performance liquid chromatography (HPLC) analysis. Fluticasone propionate content from all washings was determined by injecting duplicate 100-l samples into a reverse-phase HPLC system with a 254-nm ultraviolet detector. The HPLC mobile phase consisted of acetonitrile and water (53:47 volume:volume ratio [v:v]) with 0.03% trifluoroacetic acid. The flow rate through the system was 1.0 ml/minute. Fluticasone propionate standard curves and quality controls were prepared in a methanol and water mixture (50:50 v:v) Budesonide, as an internal standard was added to all solution before analysis. Correlation coefficients for the standard curves were 0.998.000. Assay interday coefficient of variation for fluticasone propionate quality control solutions ranged from 11.4% for the 2.0-g/ml control to 3.2% for the 20.0-g/ml control. Data Analysis The amount of fluticasone propionate deposited on each stage and throat of the cascade impactor was measured and normalized per 110 mg MDI actuation. A log probability plot constructed from these data was used to determine the distribution of particle sizes, i.e., mass median aerodynamic diameter (MMAD) and geometric standard

PAGE 39

27 deviation (GSD) [87]. The quantity of fluticasone propionate aerosol emitted within the ideal size range (1 m) for deposition and retention in the small human airways (i.e., respirable dose) and the quantity collected in the USP throat intake port (i.e., oropharyngeal dose) were also calculated. Differences among outcomes were determined by using one-way analysis of variance. The Tukey multiple comparison procedure was used to compare individual means when the overall test results were statistically significant (p<0.05). All data analysis was performed with InStat statistical software, version 3.01 (GraphPad Software Inc., San Diego, CA). Results MMAD of aerosols produced by the tested VHC devices were slightly smaller than MMAD of aerosol produced by MDI alone (p < 0.05, Table 2-1). Similarly, the tested devices reduced the GSD of aerosol compared to aerosol produced by the MDI alone (p < 0.05, Table 2-1). The differences in MMAD and GSD among the VHCs were not significant (p > 0.05, Table 2-1). Table 2-1: Mean (SD) MMAD and GSD values interpolated from the size distribution data for MDI and VHCs. MDI AeroChamber-Plus OptiChamber OptiChamber -Advantage MMAD 2.84 (0.32) 2.50 (0.05) 2.42 (0.14) 2.30 (0.06) GSD 1.26 (0.04) 1.22 (0.01) 1.22 (0.01) 1.22 (0.01) The fluticasone propionate respirable dose from all devices is shown in Figure 2-2. The mean SD fluticasone propionate respirable dose (i.e., aerosol size 1 mm) from AeroChamber-Plus (45.3 1.8 g/ actuation) and old OptiChamber (38.1 8.7 g/ actuation) were not significantly different (p > 0.05) from the respirable dose produced by MDI alone (46.7 9.7 g/ actuation). However, OptiChamber -Advantage produced

PAGE 40

28 a mean respirable dose (32.1 1.6 g/ actuation) that was significantly less than that produced by MDI alone or AeroChamber-Plus (p < 0.05) but not significantly different from old OptiChamber (p > 0.05). MDI alone Aerochamber-Plus Optichamber Optichamber-Advantage 0 10 20 30 40 50 60(46.7 9.7)(45.3 1.8)(32.1 1.6)(38.1 8.7)Respirable dose (ug/actuation) Figure 2-2. Mean SD respirable dose of fluticasone propionate from metered-dose inhaler (MDI) and valved holding chamber (VHC) devices. The fluticasone propionate oropharyngeal dose from all devices is shown in Figure 2-3. The mean SD oropharyngeal dose from MDI alone (50.5 7.2 g/ actuation) was significantly greater (p < 0.001) than the oropharyngeal dose from all three VHC devices: AeroChamber-Plus (0.9 0.3 g/ actuation), OptiChamber (0.6 0.8 g/ actuation) and OptiChamber -Advantage (0.17 0.1 g/ actuation). The difference in oropharyngeal dose between the three VHCs was not significant (p > 0.05).

PAGE 41

29 MDI alone Aerochamber-Plus Optichamber Optichamber-Advantage 0 10 20 30 40 50 60(50.5 7.2)(0.9 0.3)(0.17 0.1)(0.6 0.8)Oropharyngeal dose (ug/actuation) Figure 2-3. Mean SD oropharyngeal dose of fluticason propionate from metered-dose inhaler (MDI) and valved holding chamber (VHC) devices. Discussion and Conclusion All three VHCs tested significantly decreased the oropharyngeal dose compared with MDI alone. This reduction in oropharyngeal dose results from that large, unrespirable aerosol particles are collected in the VHC rather than in the impactor throat intake port when either VHC is attached. These data suggest that either VHC tested would reduce the risk of topical adverse effects induced by fluticasone. The highest respirable dose observed was from fluticasone propionate MDI alone. However, the respirable dose from fluticasone propionate MDI alone was not significantly larger than the respirable dose from the MDI attached to AeroChamber-Plus or OptiChamber The respirable doses from MDI alone, MDI attached to AeroChamber-Plus or OptiChamber observed in this study were not significantly different from previously reported values, respectively obtained in a nearly identical manner [93, 94]. Our results confirmed that using AeroChamber-Plus or OptiChamber

PAGE 42

30 in combination with a fluticasone propionate-110 MDI does not alter the respirable dose available to the asthmatic patients. Further, the differences in delivered particle size distribution of fluticasone aerosol, as reflected by both MMAD and GSD, between the two VHCs and MDI alone were small. This suggests that fluticasone MDI can be used successfully with AeroChamber-Plus or OptiChamber However, the respirable dose of fluticasone from OptiChamber -Advantage was 40-45% less than from the MDI alone or the AeroChamber-Plus VHC. Although the dose-response relationship in the treatment of asthma has been difficult to establish, a 40-45% decrease in respirable dose of inhaled fluticasone is likely to be a clinically relevant difference. In vitro testing has some potential limitations. The most prominent of these is that the correlation between the in vitro respirable dose of fluticasone and clinical response has not been fully evaluated. Variability in patient inhalation technique and adherence with the therapy are often the biggest determinants in clinical response and an in vitro deposition study cannot account for these variables. Future work should include measurement of aerosol deposition under more physiologic conditions likely to be encountered in asthmatic patients. Another limitation of present study is that our results are only reflective of the performance characteristics of the tested VHCs with the current fluticasone MDI formulated with a chlorofluorocarbon propellant. Our results may not be reflective of performance with other inhaled formulations of the same medication or of different ICS. Although our study was conducted at only one flow rate (28.3 L/min), other studies have shown that flow rates of 60 and 90 L/min have little effect on the respirable dose from several corticosteroids MDIs when compared with that achieved at 30 L/min [21, 95].

PAGE 43

31 There are differences in the in vitro performance characteristics of VHCs tested in our study. Our results suggest that an asthmatic patient would receive the same dose of fluticasone to the lung when using a MDI alone with proper inhalation technique or the same MDI attached to AeroChamber-Plus or OptiChamber However, the same MDI attached to OptiChamber -Advantage, a newer VHC compared with the others in this study, would result in 40-45% less fluticasone reaching the lung than that from the MDI alone. This difference is likely to be clinically relevant. Hence, in asthmatics on fluticasone MDI needing a VHC, AeroChamber-Plus or OptiChamber would be a better choice over OptiChamber -Advantage since they do not change the respirable dose compared to MDI alone.

PAGE 44

CHAPTER 3 PLASMA CONCENTRATIONS AND PHARMACOKINETICS PROFILES OF INHALED CORTICOSTEROIDS AND THEIR RELATION TO LUNG FUNCTION IN ASTHMA Introduction As we discussed in a previous chapter, inhaled corticosteroids are absorbed into the systemic circulation, predominantly from the lung and more variably from the gastrointestinal tract. Therefore they have the potential to produce adverse systemic effects. The risk of such adverse effects will depend on the extent of systemic (overall) absorption of ICS and may differ between inhaled corticosteroids due to their different physicochemical and pharmacokinetic properties. Recent studies have shown that absorption and systemic effects of one inhaled corticosteroid, fluticasone propionate, are greater in healthy volunteers than in subjects with asthma and airflow obstruction [54, 55, 96, 97]. However it was not true for budesonide [55, 98]. Another study showed that the extent of adrenal suppression following 500g fluticasone propionate was closely related to lung function with greater cortisol suppression with increasing FEV 1 [50]. Whether the systemic absorption of other inhaled corticosteroids relates to lung function is unknown. To explore this and to provide comparative data on the pharmacokinetic profiles of the four drugs in the same subjects, we compared plasma concentrations of beclometasone monopropionate, budesonide, fluticasone propionate and mometasone furoate following inhalation of a single dose in subjects with asthma who had a wide range of FEV 1 values. 32

PAGE 45

33 Methods Subjects Thirty non-smoking subjects with asthma aged between 18 and 70 were recruited between October 2003 and February 2004 from our volunteer database. Subjects were selected to provide a range of forced expiratory volume in one second (FEV 1 ) values from 30% predicted upwards. All had to have stable asthma defined as no change in asthma symptoms or treatment for at least two months. There were no restrictions based on treatment but subjects were excluded if they had significant co-morbidity, were pregnant or lactating or had a greater than 20 pack/year smoking history. The study was approved by Nottingham Research Ethics Committee and written informed consent was obtained from all subjects. Measurements Lung function FEV 1 and forced vital capacity (FVC) were measured with a dry bellows spirometer (Vitalograph, Buckingham, UK) as the higher of two successive readings within 100ml. The same spirometer was used for all measurements and was calibrated weekly. Peak expiratory flow (PEF) was measured with a mini-Wright flowmeter as the best of three readings. Plasma drug assays The plasma concentrations of budesonide, beclometasone monopropionate (the active metabolite of beclometasone dipropionate), fluticasone propionate and mometasone furoate were quantified with validated high performance liquid chromatography/tandem mass spectrometry methods using a Micromass Quattro LC-Z triple quadrupole mass spectrometer (Beverley, MA) [99-102]. The lower limits of

PAGE 46

34 detection for the assays were 15 pg/ml for beclometasone monopropionate, fluticasone propionate and mometasone furoate and 50 pg/ml for budesonide. Protocol Subjects were screened at an initial visit to measure FEV 1 assess their suitability for the study and to familiarize themselves with the placebo dry powder inhalers. Subjects taking fluticasone propionate, budesonide or mometasone furoate were changed to an equivalent dose of beclometasone dipropionate at least four days before the study day; all subjects were then asked to omit the beclometasone dipropionate from the evening before the study day to allow time for beclometasone monopropionate to be cleared from the plasma before the study. Subjects attended the department in the morning of the study day, a venous cannula was inserted and blood taken for baseline drug assay. After ten minutes rest, FEV 1 FVC and PEF were measured on two occasions five minutes apart. Subjects then inhaled two doses from each of four inhalers in random order with a mouth rinse after each drug. The drugs and total doses administered were beclometasone dipropionate 800g (Becodisks, Allen & Hanburys), budesonide 800g (Turbohaler, Astra-Zeneca), fluticasone propionate 1000g (Accuhaler, Glaxo-Wellcome) and mometasone furoate 800g (Asmanex Twisthaler, Schering-Plough). Venous blood samples were taken into sodium fluoride/EDTA tubes at intervals over the next eight hours, centrifuged at 1500 rpm for ten minutes and plasma samples extracted and frozen at -70C. The main study endpoint was the relation between the subjects % predicted FEV 1 and the area under the plasma concentration/time curve (AUC 0-8 ) for each drug. Power calculations based on a linear regression model showed that 30 patients would give 90% power to detect a correlation between FEV 1 % predicted and AUC of 0.5 or more.

PAGE 47

35 Analysis Plasma concentrations of beclometasone monopropionate, budesonide, fluticasone propionate and mometasone furoate were plotted against time for each subject; pharmacokinetic parameters (maximum plasma concentration (Cmax), time to Cmax, AUC 0-8 mean residence time and terminal half-life) were calculated using a software package (WinNonlin Professional Version 3.1 Pharsight Corporation, Mountain View, CA). Concentrations below the limit of quantification were reported as missing and extrapolated during non-compartmental analysis. Plasma drug AUC 0-8 values were plotted against the mean of the two baseline FEV 1 measurements, FEV 1 % predicted and FEV 1 /FVC ratio, and against PEF and PEF % predicted. The correlation coefficient (r) was calculated and trendlines fitted using linear regression. Partial correlation coefficients were calculated for the relationship between lung function and AUC for the four drugs after controlling for two potential confounding factors, age and gender. To estimate the magnitude of the effect of lung function on AUC values for each drug the equation derived from the linear regression model was used to predict the AUC value at 100 and 50% predicted FEV 1 and PEF and the ratio of these two values was calculated. Results All 30 subjects (mean age 57, 16 men) completed the study as per protocol. Baseline characteristics of the subjects are shown in table 3-1. FEV 1 prior to drug inhalation ranged from 36 to 138% predicted. Four subjects had detectable plasma concentrations of one inhaled corticosteroid in their baseline sample. In two subjects this was beclometasone monopropionate (1152pg/ml and 46pg/ml), in one who usually took budesonide it was budesonide (215pg/ml) and in one who usually took beclometasone

PAGE 48

36 dipropionate, fluticasone propionate was detected (19pg/ml). Data for these drugs only from the four subjects were excluded from the analysis. Table 3-1. Baseline characteristics Subject Sex Age FEV1 (liters) FEV1 (% pred) FVC (liters) FEV1/FVC PEF (liter/min) PEF (% pred) 1 M 70 1.3 48 3.6 0.35 270 50 2 M 52 3.1 90 4.3 0.71 480 79 3 M 61 1.3 36 3.9 0.32 245 42 4 F 61 2.3 98 2.9 0.79 425 97 5 F 53 2.1 86 2.8 0.74 360 79 6 M 49 2.7 78 4.2 0.65 460 75 7 M 48 2.2 66 3.7 0.59 360 79 8 M 61 3.3 106 4 0.82 555 96 9 F 45 2.7 100 3.1 0.88 425 90 10 M 61 2.5 82 3.2 0.78 450 78 11 M 55 1.4 42 3 0.47 350 58 12 M 46 3.2 74 4.3 0.74 530 82 13 F 56 3.5 138 4.1 0.86 460 101 14 F 68 1.2 62 1.9 0.63 255 62 15 F 44 3.2 113 4.3 0.74 455 95 16 F 61 1.8 72 2.15 0.81 310 70 17 F 62 2.2 96 3 0.72 465 107 18 M 50 2.6 75 4.2 0.63 430 70 19 F 35 2.4 93 3.3 0.73 460 97 20 M 67 1.9 65 3.1 0.61 390 71 21 F 58 0.9 47 2 0.44 195 45 22 M 60 1.7 53 3.2 0.53 235 40 23 M 41 2.3 57 3.8 0.59 455 71 24 F 58 2.1 87 2.6 0.78 370 83 25 F 66 1.5 77 2.3 0.66 300 72 26 M 51 1.7 46 3.2 0.54 415 67 27 M 63 1.3 43 2.8 0.46 275 48 28 F 63 2.5 99 3.2 0.8 485 110 29 F 66 0.8 44 1.5 0.55 180 43 30 M 68 1.0 40 1.5 0.55 180 43 Mean (SD) 57 (9) 2.1 (0.8) 74 (25) 3.2 (0.8) 0.6 (0.2) 376 (104) 73 (21)

PAGE 49

37 Plasma drug concentrations and pharmacokinetic profiles All four inhaled corticosteroids were present at detectable plasma concentrations following inhalation in all subjects and plasma concentrations were above the lower limit of quantification in the majority of samples (99%, 97%, 95% and 93% of samples for budesonide, beclometasone monopropionate, fluticasone propionate and mometasone furoate, respectively). 0.0 0.2 0.4 0 1 2 0 2 4 6 8 0.00 0.05 0.10 0 2 4 6 8 0.00 0.05 0.10 BMP BUD Plasma concentration (ng/ml) FP MF Time (hours) Figure 3-1. Mean (SE) plasma concentration-time curves for beclometasone monopropionate (BMP), budesonide (BUD), fluticasone propionate (FP) and mometasone furoate (MF) following inhalation (Note that Y axes have different scales) The shape of the mean plasma drug concentration-time curves and pharmacokinetic parameters varied considerably between drugs as shown in figure 3-1 and table 3-2. Data from two subjects for one inhaled corticosteroid (one beclometasone dipropionate, one

PAGE 50

38 mometasone furoate) were unsuitable for pharmacokinetic modelling and excluded from the calculations. Table 3-2. Mean (SD) pharmacokinetic parameters for beclometasone monopropionate(BMP),budesonide(BUD), fluticasone propionate(FP) and mometasone furoate (MF) BMP BUD FP MF Tmax (h) 2.93 (0.87) 0.16 (0.12) 1.19 (0.91) 1.28 (0.96) MRT (h) 12.3 (13.9) 3.6 (0.8) 11.4 (14.6) 8.4 (5.8) T 1/2 7.6 (9.7) 2.7 (0.5) 7.8 (10.0) 5.6 (4.1) Cmax(ng/ml) 0.34 (0.15) 1.59 (0.88) 0.10 (0.04) 0.09 (0.06) AUC(ng/ml/h) 1.79 (0.78) 3.43 (1.29) 0.44 (0.23) 0.43 (0.32) Peak plasma concentrations following inhalation occurred within 10 minutes with budesonide, and after 1.2, 1.3 and 2.9 hours with fluticasone propionate, mometasone furoate and beclometasone monopropionate respectively. Mean pulmonary residence times showed a similarly wide variation from 3.6 hours with budesonide to 12.3 hours with beclometasone monopropionate. The terminal half-life ranged from 2.7 hours for budesonide to 7.8 hours for fluticasone. The magnitude of the peak plasma concentration also differed markedly between drugs being highest for budesonide at 1.6 ng.ml -1 and 18, 16 and 5 times lower for mometasone furoate, fluticasone propionate and belcometasone monopropionate, respectively. AUC values ranged from 3.43 ng/ml/h for budesonide to 0.43 ng/ml/h for mometasone furoate (table 3-2). Relation between lung function and AUC AUC 0-8 values tended to be higher in patients with a higher FEV 1 % predicted for all four drugs as can be seen in figure 3-2. The correlation coefficients (r) for the relation between FEV 1 % predicted and AUC values for fluticasone propionate, budesonide, beclometasone monopropionate, and mometasone furoate were 0.30 (p = 0.11), 0.36 (p = 0.05), 0.39 (p = 0.04) and 0.55 (p = 0.002) respectively. The same pattern was seen for

PAGE 51

39 the relation between PEF % predicted and AUC values with r values for fluticasone propionate, budesonide, beclometasone monopropionate, and mometasone furoate being 0.26 (p = 0.17), 0.35 (p = 0.07), 0.42 (p = 0.03) and 0.60 (p < 0.001) respectively. The correlation between other measures of lung function (absolute FEV 1 PEF and FEV 1 /FVC ratio) and AUC was similar but generally weaker than that between FEV 1 and PEF % predicted and AUC. Both age and gender caused a 10% or greater change in the correlation coefficients. The correlation between FEV 1 % predicted and AUC was stronger after adjusting for age but weaker after adjusting for gender (Table 3-3). 0 5 10 0 5 10 0 50 100 150 0 1 2 0 50 100 150 0 1 2 BMP BUD AUC (ng.ml -1 .h) FP MF FEV 1 % predicted Figure 3-2. Relation between FEV 1 % predicted and the area under the plasma concentration-time curve (AUC) for beclometasone monopropionate (BMP), budesonide (BUD), fluticasone propionate (FP) and mometasone furoate (MF) following inhalation (Note that Y axes have different scales)

PAGE 52

40 Table 3-3. Partial correlation coefficients (r) for the relationship between FEV1 % predicated and AUC after adjusting for age or gender r p value Unadjusted 0.39 0.04 Age 0.48 0.01 Beclometasone monopropionate Gender 0.35 0.08 Unadjusted 0.36 0.05 Age 0.37 0.05 Budesonide Gender 0.28 0.15 Unadjusted 0.3 0.11 Age 0.45 0.01 Fluticasone propionate Gender 0.17 0.39 Unadjusted 0.55 0.002 Age 0.63 < 0.001 Mometasone furoate Gender 0.48 0.009 Estimated AUC values were higher at 100 % compared to 50% predicted FEV 1 for all four drugs, the ratios for budesonide, fluticasone propionate, beclometasone monopropionate and mometasone furoate being 1.3, 1.4, 1.4 and 2.3 respectively. The same pattern was seen with PEF % predicted (Table 3-4). Table 3-4. Predicted AUC values at 50 and 100% predicted FEV1 and PEF and the ratio of these values AUC at 50% predicted AUC at 100% predicted Ratio of AUC values FEV1 1.52 2.1 1.4 Beclometasone monopropionate PEF 1.45 2.22 1.5 FEV1 3.01 3.93 1.3 Budesonide PEF 2.95 4.08 1.4 FEV1 0.37 0.51 1.4 Fluticasone propionate PEF 0.37 0.51 1.4 FEV1 0.27 0.61 2.3 Mometasone furoate PEF 0.23 0.68 3 Discussion This is the first study to compare plasma concentration-time curves and pharmacokinetic profiles of beclometasone monopropionate, budesonide, fluticasone

PAGE 53

41 propionate and mometasone furoate following inhalation in the same patients and to explore the extent to which systemic absorption of all four drugs relates to lung function. Plasma Drug concentrations and Pharmacokinetic Profiles The plasma drug concentration-time curves and pharmacokinetic profiles of fluticasone propionate and mometasone furoate were broadly similar to each other but differed markedly from the other two drugs, beclometasone monopropionate and budesonide. Budesonide, the most water-soluble of the four drugs, had the shortest pulmonary residence time, time to peak plasma concentration (Tmax) and terminal half-life; beclometasone monopropionate had the longest pulmonary residence time and Tmax and its terminal half-life was also longer and similar to that of fluticasone propionate. The pharmacokinetic parameters determined for budesonide are in agreement with previous data [55, 96, 97, 103, 104] and the rapid absorption is related to its lower lipophilicity and greater solubility in bronchial fluid [105]. The shorter terminal half-life of budesonide primarily reflects its lower volume of distribution since high clearance is a feature of all the inhaled corticosteroid[106]. Budesonide had a much higher peak plasma concentration than the three other drugs, an effect attributed to high pulmonary deposition by the Turbohaler, its rapid absorption and low volume of distribution [104]. The long pulmonary residence time and Tmax with beclometasone monopropionate reflects its greater lipophilicity and greater oral bioavailability (41% compared to 11% for budesonide and <1% for fluticasone propionate and mometasone furoate [80, 106-108]) since absorption from the gastrointestinal tract is slower than that from the lungs [107]. Our values for pulmonary residence time and terminal half-life for beclometasone monopropionate were longer than those reported in studies using metered dose inhalers, [109-111] perhaps due to differences in particle characteristics and deposition pattern

PAGE 54

42 with dry powder inhalers. The lower peak plasma concentrations for beclometasone monopropionate compared to budesonide is related to its slower absorption and greater volume of distribution. Fluticasone propionate and mometasone furoate had fairly similar absorption profiles. The long terminal half-life of fluticasone propionate may be due to its large volume of distribution or flip-flop pharmacokinetics due to slow release from the lungs. Both drugs had particularly low peak plasma concentrations, even though fluticasone propionate was given at a slightly higher dose than the other three inhaled corticosteroids. Slow absorption from the lungs, increased mucociliary clearance and high volumes of distribution are likely to explain these findings. Our pharmacokinetic data for fluticasone propionate are similar to those reported previously [55, 96, 97, 112, 113], whilst the only published data on mometasone furoate are difficult to compare since a relatively insensitive assay was used [114]. Relation Between Lung Function and AUC We found that AUC values tended to be higher in patients with better lung function for all four drugs. Our estimates suggest that the systemic absorption of budesonide, fluticasone propionate, beclometasone monopropionate and mometasone furoate is some 1.3, 1.4, 1.4 and 2.3 times higher at 100% predicted FEV 1 compared to 50% predicted FEV 1. The effect appears to be greatest for mometasone furoate although the study was not powered to compare the magnitude of effect between drugs. The inhaled corticosteroids were delivered by dry powder inhalers in our study, and reduced inspiratory flow may have caused less drug to be delivered to the airways in subjects with greater airflow obstruction. Furthermore, the drug that enters the airways is more likely to be deposited in central airways in patients with airflow obstruction, and

PAGE 55

43 hence more likely to be removed by mucociliary clearance. This would be expected to affect drugs that dissolve slowly in airway lining fluid such as mometasone furoate and fluticasone propionate most and drugs with high water solubility, such as budesonide, least [105]. When we explored the effect of age and gender on the relation between lung function and AUC values we found the relation to be stronger after adjusting for age but weaker after adjusting for gender. This highlights the fact that factors other than lung function affect the systemic absorption of inhaled corticosteroids thus reducing the strength of the relation between lung function and plasma concentration AUC values in cross-sectional studies. This conclusion is supported by our later study, presented in next chapter, in which we compared plasma concentrations of budesonide and fluticasone propionate following inhalation with and without prior methacholine-induced bronchoconstriction (mean fall in FEV 1 of 33%). Mean AUC values were reduced for both drugs after lowering FEV 1 i.e. when we looked at within subject changes, but not when we compared differences in FEV 1 on AUC between subjects. Studies of the effect of lung function on the systemic absorption of inhaled corticosteroids across populations need to allow for between-subject variability. Conclusion In conclusion our study has shown marked differences in plasma concentration-time curves and pharmacokinetic profiles between beclometasone monopropionate, budesonide, fluticasone propionate and mometasone furoate following inhalation. The net effect of these differences on the benefit to risk ratio of inhaled corticosteroids is complex and needs to be explored further. The systemic absorption of all four inhaled corticosteroids appears to be greater in patients with a higher FEV 1 % predicted although

PAGE 56

44 other factors, including age and gender affect this relationship. Our findings re-enforce the importance of reviewing the need for higher doses of an inhaled corticosteroid, particularly in patients with relatively normal lung function and in patients whose lung function improves with treatment.

PAGE 57

CHAPTER 4 FLUTICASONE AND BUDESONIDE CONCENTRATIONS AFTER INHALATION EFFECT OF BRONCHOCONSTRICTION Introduction Adverse systemic effects from inhaled corticosteroids have been recognized increasingly over recent years. With around 5 % of the population in more economically developed countries prescribed an inhaled corticosteroid, often for many decades, understanding the factors that determine these adverse effects is important. All currently available inhaled corticosteroids are absorbed into the systemic circulation and hence have the potential to cause adverse systemic effects. The risk may differ between inhaled corticosteroids, however, due to their different physicochemical and pharmacokinetic properties and there is evidence that these drug related differences interact with patient factors such as airflow obstruction [115]. Plasma drug concentrations following inhalation of 1000 g fluticasone were considerably lower in people with airflow obstruction than in healthy volunteers but this was not the case for budesonide [51, 55, 97]. The differences between fluticasone and budesonide were attributed to differences in lipophilicity between the two drugs [55]. Whether changes in airflow obstruction within an individual affect systemic absorption in the same way is unknown, but important in view of the fluctuations in airway caliber seen in asthma. To explore this we compared the plasma concentrations of fluticasone and budesonide following inhalation of a single dose, with and without prior methacholine-induced bronchoconstriction. 45

PAGE 58

46 Methods Subjects Twenty non-smoking subjects with asthma aged between 18 and 70 were recruited from our volunteer database. To be included subjects had to have a forced expiratory volume in one second (FEV 1 ) of at least 80 % predicted and 1.5 litres, a provocative dose of methacholine causing a 20 % fall in FEV 1 (PD 20 ) below 8 M, and stable asthma, defined as no change in asthma symptoms or treatment for two months. Subjects were excluded if they had significant co-morbidity, were taking any medication known to alter the metabolism of corticosteroids, were pregnant or lactating, or had a greater than 20 pack year smoking history. The study was approved by Nottingham Research Ethics Committee and written informed consent was obtained from all subjects. Measurements Spirometry FEV1 and forced vital capacity (FVC) were measured with a dry bellows spirometer (Vitalograph, Buckingham, UK) as the higher of two successive readings within 100 ml. Methacholine inhalation challeng e Subjects inhaled three puffs of normal saline followed by doubling doses of methacholine from 0.048 M to a maximum of 24.5 M from DeVilbiss nebulizers and FEV1 was measured one minute after each dose [116] The test was stopped once FEV1 had fallen by 20 % from the post-saline value during screening, for calculation of PD20 by interpolation, and once FEV1 had fallen by 25 % in the main study.

PAGE 59

47 Drug assays Plasma concentrations of fluticasone propionate and budesonide were quantified with previously validated high performance liquid chromatography/tandem mass spectrometry methods using a Micromass Quattro LC-A triple quadrupole mass spectrometer (Beverley, MA) [99, 100]. The lower limits of detection for the assays were 15 pg/ml for fluticasone propionate and 50 pg/ml for budesonide. The intraand inter-assay coefficients of variation for the assays were below 13.6 for both fluticasone and budesonide at concentrations between 0.015 and 0.75 ng/ml for fluticasone and between 0.15 and 2.5 ng/ml for budesonide. Protocol Subjects were screened to assess suitability and to measure FEV 1 and PD 20 methacholine. Subjects familiarized themselves with the two dry powder inhalers (Accuhaler, Glaxo-Wellcome and Turbohaler, Astra-Zeneca) to ensure optimal inhaler technique during the study and those taking fluticasone or budesonide were changed to an equivalent dose of beclometasone dipropionate for four days before and until study completion. Subjects were asked to avoid short and long acting 2 agonists for 12 hours, and to avoid exercise and caffeine on the morning of the study. Subjects attended for two study visits at the same time of day one hour, 7 3 days apart. A venous cannula was inserted and after ten minutes rest, FEV 1 was measured. 1000 g fluticasone (Accuhaler), given as two 500 g doses, and 800 g budesonide (Turbohaler), given as two 400 g doses, were then inhaled in random order; after each drug subjects rinsed their mouths with water which was then discarded. Venous blood samples were taken into heparinised tubes at intervals over the next 5 hours (0, 5, 15, 30, 60, and 90 minutes, 2, 3, 4 and 5 hours), centrifuged at 1500 rpm for 10 minutes and

PAGE 60

48 plasma samples were then frozen at C. The protocol for the two study visits was identical except that on one occasion prior to inhalation of the drugs, a methacholine challenge was carried out to induce a fall in FEV 1 of at least 25 % from baseline FEV 1 The two studies were carried out in random order according to a computer generated code. Power calculations showed that 20 patients would give 90 % power to detect a 0.7 SD difference in the area under the plasma concentration time curve over five hours, the primary outcome, between the two study visits. Analysis Plasma fluticasone and budesonide concentrations were plotted against time for each subject and the maximum plasma concentration (C max ), time to maximum plasma concentration (T max ) and area under the curve up to five hours (AUC 0-5 ) calculated using a software program (WinNonlin Professional Version 3.1, Pharsight Corporation, Mountain View, CA). C max T max and AUC 0-5 values for the two study visits were compared for each drug by paired t-test. To allow for the differences in concentrations of the two drugs we calculated the ratio of the AUC 0-5 values on the days with and without bronchoconstriction for each subject for each drug and compared the values between drugs by Wilcoxon signed rank test. This analysis was repeated for C max values. Results All 20 subjects (mean age 48, 12 men) completed the study and their baseline characteristics are shown in table 4-1. All were non-smokers and only two had ever smoked. Subjects had a geometric mean PD 20 methacholine of 1.2 M (range 0.06 to 6.1). Mean (SD) FEV 1 was 2.9 (0.7) litres (91 % predicted) on screening and 2.8 (0.7) and 2.9 (0.6) litres on arrival for the visits with and without bronchoconstriction. All subjects had

PAGE 61

49 a fall in FEV 1 of at least 25 % (range 25 to 47 %) following methacholine giving an FEV 1 of 1.9 (0.5) litres (60 % predicted) prior to drug inhalation and a mean difference in FEV 1 prior to drug inhalation on the two days of 0.9 litres (range 0.5 to 1.35) or 33 % (range 20 to 45 %). Table 4-1. Baseline characteristics Subject Sex Age Height (cm) Weight (kg) FEV1 (liters) FEV1 % pred PD20 (micromol) Usual ICS 1 M 55 192 78 3.8 91 2.7 BDP* 2 M 50 173 78 2.8 80 3.4 None 3 F 50 160 86 2.55 103 5.7 BDP 4 M 63 174 77 2.8 89 2.7 BDP 5 F 67 155 68 1.55 84 0.1 None 6 M 23 170 79 4 96 2.3 None 7 F 39 161 63 3.5 126 0.5 BDP 8 M 38 189 92 3.8 84 0.9 None 9 M 41 182 96 3.8 92 0.4 BUD 10 M 41 176 87 3.8 98 5 None 11 M 52 172 88 3.05 90 0.1 None 12 F 62 166 105 1.95 81 0.06 BDP 13 M 66 173 78 2.4 80 6.1 BDP 14 M 48 166 79 2.75 84 1 FP 15 M 49 164 68 2.8 89 5.1 BUD 16 M 42 179 67 3.55 89 0.4 BDP 17 F 65 161 64 2 93 6.1 None 18 F 27 157 57 2.8 96 2.2 FP 19 F 34 161 71 2.9 100 1.1 FP 20 F 44 166 60 2.35 82 0.6 FP Mean (SD) 48 (13) 167 (10) 77 (13) 2.9 (0.7) 91 (11) 1.2** Beclometasone dipropionate ** Geometric mean Data for budesonide plasma concentrations were lost for two subjects due to a computer problem during drug analysis. Complete data were therefore available for 20 subjects for fluticasone and 18 subjects for budesonide.

PAGE 62

50 The shape of the fluticasone and budesonide plasma concentration-time curves showed marked differences on the study day without bronchoconstriction. Peak plasma concentrations of budesonide were 14 times higher than those of fluticasone and occurred considerably earlier as shown in figure 4-1; mean (SD) time to maximum concentration (T max ) values were 0.21 (0.24) hr and 1.21 (0.91) hr for budesonide and fluticasone respectively. The shape and time course of the plasma drug concentration curves for each drug were similar on the two study days although plasma levels were lower following methacholine-induced bronchoconstriction (figure 4-1 and table 4-2). Table 4-2. Pharmacokinetic parameters for fluticasone and budesonide when inhaled with and without prior methacholine-induced bronchoconstriction Fluticasone Budesonide Without bronchoconstriction (SD) Following bronchoconstriction (SD) Without bronchoconstriction (SD) Following bronchoconstriction (SD) Cmax (ng.ml -1 ) 0.12 (0.07) 0.06 (0.02) 1.67 (0.82) 1.16 (0.48) Tmax (h) 1.21 (0.91) 0.78 (0.62) 0.21 (0.24) 0.23 (0.24) MRT (h) 7.79 (8.13) 4.88 (1.79) 3.09 (1.09) 3.17 (1.73) Cmax Maximum plasma concentration Tmax Time of maximum plasma concentration MRT Medium residence time Mean C max values for both fluticasone and budesonide were lower when the drugs were inhaled following bronchoconstriction (table 4-2). C max values were a median 44 % lower (interquartile range (IQR) 21 to 60) for fluticasone and 33 % lower (IQR -7 to 51) for budesonide following bronchoconstriction but the ratio between study visits for fluticasone was not significantly different to the ratio for budesonide, p=0.18. Mean AUC 0-5 values, the primary endpoints, for both fluticasone and budesonide were lower when the drugs were inhaled following bronchoconstriction (table 4-3). AUC 0-5 values were a median 60 % lower (IQR 36 to 75) for fluticasone and 29 % lower (IQR 2 to 44) for budesonide following bronchoconstriction. The ratio for AUC 0-5 values

PAGE 63

51 between study visits for fluticasone was significantly greater than the ratio for budesonide, p=0.007. 0.00 0.05 0.10 0.15 0.20 Plasma concentration ( n g /ml ) 0 1 2 3 4 5 0.00 0.50 1.00 1.50 2.00 Figure 4-1. Mean (SE) plasma drug concentrations following inhalation of 1000 g fluticasone (Accuhaler) and 800 g budesonide (Turbohaler) in 20 subjects with asthma with ( ) and without ( ) prior methacholine-induced bronchoconstriction (Note: Y axes have different scales). Discussion This is the first study to explore the effect of change in airflow obstruction within an individual on plasma concentrations of fluticasone and budesonide following drug

PAGE 64

52 inhalation. There was an average 33 % difference in FEV 1 prior to drug inhalation on the two study visits as a result of methacholine challenge. Plasma concentrations of both fluticasone and budesonide were lower when FEV 1 was reduced but the magnitude of this effect was greater for fluticasone. Table 4-3. Area under the curve (AUC) values for fluticasone and budesonide when inhaled with and without prior methacholine-induced bronchoconstriction Fluticasone Budesonide Without broncho-constriction (SD) Following broncho-constriction (SD) Differences (95% CI), p value Without broncho-constriction (SD) Following broncho-constriction (SD) Differences (95% CI), p value AUC (ng.ml -1 .h) (SD) 0.40 (0.23) 0.16 (0.10) 0.23 (0.14 to 0.33), p <0.001 2.87 (1.19) 2.28 (0.90) 0.60 (0.02 to 1.17), p =0.04 The greater effect of bronchoconstriction on plasma concentrations of fluticasone compared to budesonide is similar to the findings in patients with and without airflow obstruction [55, 96, 97, 105]. Bronchoconstriction will cause a greater proportion of both drugs to be deposited in more central airways, where mucociliary clearance is better able to remove drug [117]. This would be expected to affect fluticasone more than budesonide because fluticasone, being more lipophilic, dissolves more slowly in airway lining fluid and hence is available for clearance for a longer time [105]. Furthermore little of the fluticasone that is removed by mucociliary clearance will reach the systemic circulation due to its low oral bioavailability (<1 % compared to around 10 % for budesonide) [105]. Peak plasma drug concentrations occurred earlier and were considerably higher following inhalation of budesonide than fluticasone in keeping with previous data [55]. These differences are likely to reflect high pulmonary deposition of budesonide from the Turbohaler, its higher oral bioavailability, and greater water solubility and its lower volume of distribution [105, 115, 118].

PAGE 65

53 We gave fluticasone and budesonide at the same time to ensure that they were studied under identical conditions. This approach was validated by Agertoft and Pederson who found similar plasma fluticasone and budesonide concentrations following inhalation whether the drugs were given separately or together [119]. Drug concentrations were measured over five hours since our previous study showed that the major differences in plasma concentrations between subjects with and without airflow obstruction occurred during this time [55]. We did not seek to obtain pharmacokinetic parameters other than C max and T max since these are already available and a longer sampling period would have been required [115]. We believe our findings are due to reduced rather than delayed, drug absorption in view of the shape of the plasma concentration-time curves and the fact that the changes within subjects in this study are very similar to the differences seen over 8 and 12 hours between subjects with and without airflow obstruction[51, 55, 97]. By using a methacholine challenge we were able to study the effect of change in lung function within an individual on plasma concentrations of fluticasone and budesonide under controlled conditions. This is likely to be a reasonable model for changes in airflow that occur as asthma control deteriorates. A 33 % reduction in FEV 1 in our study caused a 60 % reduction in AUC 0-5 values for fluticasone suggesting a 60 % reduction in systemic exposure. Equally, as FEV 1 increases with treatment, systemic exposure would be expected to increase in a similar way. The risk of adverse systemic effects from inhaled fluticasone is likely, therefore, to vary considerably in relation to lung function within an individual while that from budesonide is unlikely to be affected to the same extent. These findings re-enforce the importance of reviewing the need for

PAGE 66

54 higher doses of inhaled corticosteroids and particularly fluticasone, as lung function improves and in patients with relatively normal lung function.

PAGE 67

CHAPTER 5 EVALUATION OF THE ADMINISTRATION TIME EFFECT ON THE CUMULATIVE CORTISOL SUPPRESSION AND CUMULATIVE LYMPHOCYTES SUPPRESSION FOR ONCE-DAILY INHALED CORTICOSTEROIDS Introduction Inhaled corticosteroids (ICS) continue to remain the first-line therapy for the treatment of asthma [12]. Clinical studies have demonstrated the high anti-asthmatic efficacy (e.g. prevention and control of symptoms) and reduced systemic side effects of ICS therapy. However, potential systemic side effects induced by orally (swallowed) or pulmonary absorbed drug remain a concern in the eyes of the patient [13, 64, 69]. The major systemic effects include bone demineralization, growth suppression in children, hypothalamic-pituitary-adrenal (HPA) axis suppression and etc sec. Since many of these effects are either difficult to quantify or not possible to measure during short-term observations, the use of biomarkers is common in clinical pharmacology studies of corticosteroids. Two such markers have been identified for the systemic effects of corticosteroids: suppression of endogenous cortisol and blood lymphocytes [120, 121]. Cortisol serum levels are very sensitive to the presence of exogenous corticosteroids and the release of serum cortisol is often dramatically suppressed after drug administration. Thus, cortisol suppression as monitored by serum or urinary cortisol is currently used in clinical pharmacology studies for the assessment of systemic side effects, as these parameters are sensitive and relatively easy to determine while other more clinically relevant parameters (e.g. growth suppression) is used less frequently in specific patient populations. Twenty-four hour cortisol suppression has been used as the 55

PAGE 68

56 most sensitive marker [122] and adequate PK/PD models have been developed [123]. Endogenous serum cortisol concentrations follow a circadian rhythm. Peak concentration is reached in the morning between 6AM and 10 AM, trough concentrations are reached at night between 8PM and 2AM [124-126]. Therefore, cortisol suppression after ICS is expected to undergo diurnal variation dependent on the administration time in the day [127]. It is known that the administration of corticosteroids results in a transient depletion of blood lymphocytes, which is called lymphocytopenia [128, 129]. Lymphocytopenia seems to be an appropriate choice as biomarker since the decline in lymphocytes is directly related to the suppression of immune system often observed in chronic corticosteroid therapy, whether it is side effect or desired outcome [130-134]. Endogenous cortisol, like exogenous corticosteroids, also affects the number of lymphocytes in the blood. Since the endogenous serum cortisol concentration follows a circadian rhythm, in the absence of additional stimuli the lymphocytes count shows intraday fluctuations inversely related to cortisol concentrations [121]. The overall extent of lymphocytopenia observed after exogenous corticosteroid administration is the combined effect of a direct suppression of blood lymphocytes by the exogenous glucocorticoid and a reduction of this effect due to the decrease in serum cortisol level induced by exogenous glucocorticoids. Therefore, the extent of lymphocytes suppression in the blood after ICS administration may also undergo diurnal variation dependent on the administration time of the ICS during the day. The importance of the administration time of corticosteroids for lymphocytes suppression has not yet been recognized.

PAGE 69

57 Newer inhaled glucocorticoids are currently being dosed once a day. Clinical studies in recent years have tried to evaluate the effect of morning or evening dosing on the effect and systemic side effect with contradicting results [41-48] In this study, a population modeling/simulation approach was developed and applied to evaluate the effect of daily administration time on 24-hour cumulative cortisol suppression (CCS) and cumulative lymphocytes suppression (CLS). Methods Historic Data Results of a previously published study comparing the single-dose and steady-state (day 5) pharmacokinetics and pharmacodynamics of inhaled fluticasone propionate and budesonide in healthy subjects were used to develop the necessary population pharmacokinetic and pharmacodynamic model [103]. The study was conducted in accordance with the revised Declaration of Helsinki (Hongkong Revision 1989) and to Good Clinical Practice guidelines. A total of 14 healthy male volunteers, at mean age of 26.4 years (range 22-32), height of 179.2 (range 172-187) cm and weight of 72.7 (range 62-85) kg completed the study. The study evaluated the single dose and steady state pharmacokinetic and pharmacodynamic properties for inhaled fluticasone propionate (200 and 500 g via Diskus FP-DSK) and budesonide (400 and 1000 g via Turbohaler BUD-TBH) in healthy subjects. During this double blind, double dummy, randomized, placebo controlled, five-way crossover study single doses were administered at 8 am on day 1 followed by twice-daily dosing at 8 a.m. and 8 p.m. on days 2-5. During day 1 and 5 blood samples for drug, cortisol and lymphocytes analysis were collected at

PAGE 70

58 frequent intervals. On days 2-4, only trough samples (8 a.m. and 8 p.m.) were obtained. The detail of the study has been reported elsewhere [103]. Population Pharmacokinetic/Pharmacodynamic Analysis A sequential modeling approach was taken to facilitate the data analysis. In the first step, the population pharmacokinetic analysis was performed. This was followed by the population pharmacodynamic analysis with individual pharmacokinetic parameter estimates as part of input. Log-transformed data were used in the analysis since it was found that log-transformation of the data stabilized the model. The analysis was carried out using the NONMEM program (Version V, Globomax, Hanover, Md). Pharmacokinetic analysis Based on previously performed standard two-stage pharmacokinetic analysis, 1 and 2 compartment body model with first-order absorption were both tried to identify the optima structure model for BUD and FP, respectively [135, 136]. The inter-subject variability of PK parameters was modeled as exponential error model, as follows: iiP exp where P i is the parameter for ith subject, is the typical population value of the parameter, and i is a random inter-subject effect with mean 0 and variance 2 The intra-subject variability was modeled as additive error model, as follows: ijpijijyy lnln where y ij and y pij represent the ith subjects jth observed and predicted concentration, respectively, and is the random residual error, which is normally distributed with mean 0 and variance 2 Fist-order conditional method (FOCE) was used for parameter estimation. Individual subject PK parameters were calculated using the posterior conditional estimation technique of NONMEM. After determination of structure

PAGE 71

59 model, inter-occasion variability was explored for all PK parameters [137]. The significance level required to keep inter-occasion variability on a parameter in a hierarchical model was set to p < 0.01, which corresponds to a drop in the objective function by 6.64. Finally, a bootstrap resampling technique was applied as internal validation of the model [138]. Wings for NONMEM (Version 3.04, developed by Dr. N. Holford) was used for execution of bootstrap resampling analysis. Pharmacokinetic/Pharmacodynamic analysis of effects of BUD and FP on cortisol In the absence of exogenous stimuli, the circadian rhythm of endogenous cortisol was describe by a linear release model, as follows [139]: RRVdtttRtVdttC maxmaxminmaxminmaxmin()(2424 ) t max < t < t min RRVdtttRtVdttC maxmaxminmaxminmaxmin()( ) t min < t < t max CorttotCorteCCorttotCkRdtdC where Rc is the release rate of cortisol, Rmax is the maximum cortisol release rate, is the time of maximum release, is the time of minimum release of cortisol, Vd is the volume of distribution of cortisol and fixed at literature value 33.7 L, C tot Cort is the cortisol concentration, Ke is the elimination rate constant of cortisol and fixed at literature value 0.56 h -1 The change in total cortisol plasma concentration and its modulation by the free plasma concentration of an exogenous corticosteroid was described with an indirect response model, as follows [140]:

PAGE 72

60 CorttotCorteECSfCortECSECSfCCorttotCkCECCERdtdC50max1 where E max is the maximum suppressive effect and fixed at 1, C f ECS is the unbound exogenous corticosteroid concentration and E 50 ECSCort is the unbound exogenous corticosteroid concentration to achieve 50% of the maximum suppression [139, 140]. The unbound faction in plasma is 0.12 for BUD, and 0.1 for FP [80]. The inter-subject variability of PD parameters was modeled as exponential error model, as described in previous section, and residual error was modeled as additive error model, as described in previous section. Fist-order (FO) method was used for parameter estimation. Individual subject PD parameters were obtained as empirical Bayes estimates using the POSTHOC option of NONMEM. The cortisol data after placebo were used to calculate t max and t min The cortisol data after active treatment were modeled with PK parameters, t max and t min fixed to estimate Rmax and EC 50 ECSCort Inter-occasion variability was explored for all PD parameters. The significance level required to keep inter-occasion variability on a parameter in a hierarchical model was set to p < 0.01, which corresponds to a drop in the objective function by 6.64. The final model was validated by prediction check: the serum cortisol concentrations were simulated 100 times, then 24-hour CCS at steady state and observed 24-hour CCS at steady state were compared using Kruskal-Wallis test followed by Dunns multiple comparison. The 24-hour CCS is quantified as follows [127]: 24,24,24,24,24,%BaseTherapyBaseBaseSuppAU C AUCAUCAU C AUCCCS

PAGE 73

61 where AUC Supp,24 is the difference between the area under the cortisol plasma concentration-time curve at baseline (AUC base,24 ) and during therapy (AUC therapy,24 ) over 24 hour. Pharmacokinetic/Pharmacodynamic analysis of effects of budesonide and fluticasone propionate on lymphocytes The combined pharmacodynamic effects of cortisol and exogenous corticosteroid on lymphocyte can be described as [141]: NkCECECCECCECECCEKdtdNoutCortfLymCortLymECSECSfLymECSCortfLymCortLymECSECSfmaxin50505050501 where N is the number of lymphocytes in blood, CfCort is the unbound cortisol concentration, Kin is the influx rate constant of cells from the extravascular space, Kout is and fixed at 1, E50 is the unbound concentration of cortisol to produce 50% of the maximum effect and E50E is the unbound concentration of to achieve 50% of the maximum effect. Cortisol exhibits non-linear binding to serum protein in the solving following equation [120]: the efflux rate constant of cells out of the vascular space, Emax is the maximum effect CortLymCSLymphysiological concentration range and the unbound concentration can be calculated by Cort CortfCortfAlbAlbCortfTc fTcTcCorttotCCQKCKCQKC1 where KTc is the affinity constant between cortisol and transcortin (3107 1/mol), KAlb the affinity constant between cortisol and albumin (5000 l/mol), QTc the total transcortin concentration (0.7 mol/l), QAlb the total albumin concentration (550 mol/l) and Cfrt the unbound cortisol serum concentration [142]. Co

PAGE 74

62 The inter-subject variability of PD parameters was modeled as exponential error model, and residual error was modeled as additive error model. Fist-order (FO) method was ur asion variability on a parameter in a hierarchical model was se sed for parameter estimation. Individual subject PD parameters were obtained as empirical Bayes estimates using the POSTHOC option of NONMEM. The lymphocyte data were modeled to estimate K in K out EC 50 CortLym and EC 50 ECSLym with all the othePK and PD parameters fixed. Inter-occasion variability was explored for all PD parameters. The significance level required to keep inter-occ et to p < 0.01, which corresponds to a drop in the objective function by 6.64. Thfinal model was validated by prediction check: the blood lymphocyte counts were simulated 100 times, then 24-hour CLS at steady state and observed 24-hour CLS at steady state were compared using Kruskal-Wallis test followed by Dunns multiplecomparison. The 24-hour CLS is quantified as follows [127]: 24, 24,24,24,%TherapyBaseSuppAU 24,BaseBase C AUCAUCAU C AUCCLS where AUCSupp,24 is the difference between the area under the lymphocytes plasma number of counts time curve at baseline (AUCbase,24) and during therapy (AUCtherapy,24) CLS for Once-daily Dose of BUD and FP Using the developed population PK/PD model, we performed simulations to evect on 24-hour CCS and 24-hour CLS at steadyceover 24hr. Population PK/PD Simulation to Evaluate Administration Time Effect on CCS and aluate the effect of the administration time eff state for 4 clinically relevant dosing regimens: 1mg once-daily BUD, 2mg ondaily BUD, 0.5mg once-daily FP and 1mg once-daily FP. For each dosing regimen,

PAGE 75

63 simulation was performed at 24-hourly distributed administration time points throughouthe day. NONEME was used for execution of the simulation.75 hypothetic subjects aincluded in the simulation. The 75 subjects parameters were assumed to follow the same distribution as the 14 subjects whose population means and variances were estimated inabove analysis. Each subjects parameters were randomly sampled from the estimated populations respectively, and kept the same throughout the simulation. For those parameters whose inter-subject variability was not estimated, the values of population mean, either estimated or fixed in above analysis, were applied to each subject in tsimulation. The whole simulation for the four dosing regimens was repeated 8 times widifferent seed numbers assigned in NONMEM, which were to generate different grouof 75 hypothetic subjects in each simulation. For each simulation, the 75 subjects 24-hour CCS at steady state resulted from different administration time were calculated and then compared, respectively for each dosing regimen, using Friedman test followed byDunns multiple comparison. Similarly, the 75 subjects 24-hour CLS at steady state resulted from different administration time were calculated and then compared, respectively for each dosing regimen in each simulation, using Friedman test followedDunns multiple comparison. Results Population Pharmacokinetic t re he th ps by Model Budesonide osen as structure model for BUD since two-model was over-parameterized and one-compartment model was sufficient to des A one-compartment model was ch compartment cribe the data (Appendix A for NONMEM code). No trend was observed in thediagnostic plot of weighted residuals versus time, indicating no bias in the model

PAGE 76

64 prediction. The relationship between population predicted, individual predicted and observed plasma concentrations as a function of time is shown in Figure 5-1. Therno significant inter-occasion variability found for any parameter (Table 5-1) based onlikelihood ratio test (P<0.01, e was 2 0.05 = 6.64, df = 1). The parameter estimates were in good agreement with the mean parameter estimates of the bootstrap replicates (Table 5-1), indicating stability of the model [143]. The value for the absorption rate (Ka) was fixed because BUDs fast absorption and plasma sampling time scheme did not allow estimation of this parameter. 020406080100120Time (hr) -3-1 1 log (Pop. Predi and Observed DV, ng/ml)log (Pop. Predi and Observed DV, ng/ml)log (Pop. Predi and Observed DV, ng/ml)log (Pop. Predi and Observed DV, ng/ml) cted DVcted DVcted DVcted DV 020406080100120Time (hr) -3-11 log (Individual Predicted DV and Observed DV, ng/ml) Figure 5-1. Relationship between population predicted (open triangle) and observed BUD (open circle) concentrations, individual predicted (open square) and observed BUD (open circle) concentrations as function of time.

PAGE 77

65 Table 5-1. Summary of BUD pharmacokinetic model parameter estimates Parameters Original dataset estimate (RSE %) 200 bootstrp replicates mean (RSE %) Ka h -1 10 (Fixed) 10 (Fixed) K10 h -1 0.21 (5.5) 0.21 (5.3) V L 1320 (9.8) 1336 (9.2) Ka %CV K10 %CV 15.5 (85.4) 14.2 (46.5) V %CV 31.2 (37.2) 30.0 (20.6) ng/ml 0.54 (17.5) 0.54 (8.7) RES, relative standard error; CV, coefficient of variation; Ka, absorption rate; K10, elimination rate constant; V, volume of distribution central compartment; = not calculated. Fluticasone propionate A two-compartment model was chosen as structure model for FP since one-compartment model did not fit data well (Appendix B for NONMEM code). No trend was observed in the diagnostic plot of weighted residuals versus time. The relationship between population predicted, individual predicted and observed plasma concentrations as a function of time is shown in Figure 5-2. There was no significant inter-occasion variability found on any parameter (Table 5-2) based on likelihood ratio test (P<0.01, 2 0.05 = 6.64, df = 1). The parameter estimates were in good agreement with the mean parameter estimates of the bootstrap replicates (Table 5-2), indicating stability of the model. Population Pharmacokinetic/Pharmacodynamic Model The population PK/PD model provided reasonable fits for the cortisol and lymphocyte data (Appendix C and D for NONMEM code). No trend was observed in the diagnostic plot of weighted residuals versus time. The relationship between population predicted, individual predicted and observed plasma concentrations as a function of time is shown in Figure 5-3, and 5-4. The parameter estimates are listed in Table 5-3. A

PAGE 78

66 significant inter-occasion variability was found for Rmax for both BUD and FP. None of the 24-hour CCS and CLS at steady state calculated from simulated cortisol concentrations and lymphocyte counts time-profile was different from observed CCS and CLS in the original study [112], respectively (p < 0.05), indicating stability of the model. 020406080100120Time (hr) -5.5-4.0-2.5 log (Pop. Predicted DV and Observed DV, ng/ml)log (Pop. Predicted DV and Observed DV, ng/ml)log (Pop. Predicted DV and Observed DV, ng/ml)log (Pop. Predicted DV and Observed DV, ng/ml) 020406080100120Time (hr) -5.5-4.0-2.5 log (Individual Predicted DV and Observed DV, ng/ml) Figure 5-2. Relationship between population predicted (open triangle) and observed FP (open circle) concentrations, individual predicted (open square) and observed FP (open circle) concentrations as function of time. Population Pharmacokinetic/Pharmacodynamic Simulation The relationship between administration time and corresponding mean CCS and CLS of the 75 subjects from each simulation is shown in Figure 5-5 and 5-6 (Appendix E and F for NONMEM code). For once-daily of BUD, maximum CCS was observed when BUD was given in 2am-5am, while minimum CCS was observed when BUD was given

PAGE 79

67 in 3pm-6pm. Similarly, for once-daily of FP, maximum CCS was observed when FP was given in 4am-8am, and minimum CCS was observed when FP was given in 3pm-8pm. On the contrary to what we observed for CCS, maximum CLS was observed when BUD was given in 3pm-8pm and when FP was given in 3pm-7pm, while minimum CLS was observed when BUD or FP was given in 3am-9am. The maximum and minimum mean CCS and CLS of the subjects and the corresponding administration time points of the 8 simulations are summarized in Table 5-4 and 5-5. The difference between the maximum and minimum mean suppression of the subjects are also summarized in Table 5-4 and 5-5. The maximum mean CCS of the subjects is significantly different from minimum mean CCS of the subjects, respectively for each dosing regimen of each simulation (p < 0.001). In addition, the maximum mean CLS of the subjects is significantly different from minimum mean CLS of the subjects, respectively for each dosing regimen of each simulation (p < 0.001). Table 5-2. Summary of FP pharmacokinetic model parameter estimates Parameters Original dataset estimate (RSE %) 200 bootstrp replicates mean (RSE %) Ka h -1 2.79 (13.7) 3.3 (39.9) K12 h -1 0.21 (13.0) 0.24 (56.7) K21 h -1 0.16 (8.0) 0.18 (26.6) K10 h -1 0.14 (6.9) 0.13 (11.7) V L 4130 (8.0) 4454 (18.2) Ka %CV 13.7 (95.2) 11.4 (60.8) K12 %CV 16.1 (100.8) 18.6 (72.4) K21 %CV 30.1 (67.1) 28.5 (77.9) K10 %CV 14.5 (70.3) 14.4 (96.3) V %CV 10.7 (80.6) 9.2 (49.6) ng/ml 0.27 (10.0) 0.26 (5.5) RES, relative standard error; CV, coefficient of variation; Ka, absorption rate; K12, transfer rate constant from central compartment to peripheral compartment; K21, transfer rate constant from peripheral compartment to central compartment ; K10, elimination rate constant; V, volume of distribution central compartment; = not calculated.

PAGE 80

68 020406080100120Time (hr) 0100200 Individual Predicted DV and Observed DV (ng/ml) A 020406080100120Time (hr) 5090130 Individual Predicted DV and Observed DV (%) B Figure 5-3. Diagnostic plots of BUD PK/PD model. (a) Relationship between individual predicted (open square) and observed (open circle) cortisol concentrations as function of time for BUD PK/PD model. (b) Relationship between individual predicted (open square) and observed (open circle) lymphocyte counts as function of time for BUD PK/PD model.

PAGE 81

69 020406080100120Time (hr) 0100200 Individual Predicted DV and Observed DV (ng/ml) A 020406080100120Time (hr) 5090130 Individual Predicted DV and Observed DV (%) B Figure 5-4. Diagnostic plots of FP PK/PD model. (a) Relationship between individual predicted (open square) and observed (open circle) cortisol concentrations as function of time for FP PK/PD model. (b) Relationship between individual predicted (open square) and observed (open circle) lymphocyte counts as function of time for FP PK/PD model.

PAGE 82

70 Table 5-3. Summary of FP and BUD pharmacokinetic/pharmacodynamic model parameter estimates BUD FP Parameters Mean (RSE %) Mean (RSE %) Tmax h 22.7 (2.5) 22.7 (2.5) Tmin h 17.5 (0.7) 17.5 (0.7) Vd L 33.7 (Fixed) 33.7 (Fixed) Ke h -1 0.56 (Fixed) 0.56 (Fixed) Rmax ng/h 3270 (4.1) 2490 (7.6) EC50 ICScort ng/ml 0.02 (20.9) 0.008 (37.7) Kin %/h 21.7 (25.3) 14.4 (60.8) Kout, %/h 0.19 (28.4) 0.13 (77.5) EC50 Cortlym ng/ml 30.8 (15.6) 26.8 (31.8) EC50 ICSlym ng/ml 0.16 (36.9) 0.05 (89.1) Tmax %CV 7.0 (54.1) 7.0 (54.1) Tmin %CV 4.0 (33.9) 4.0 (33.9) Vd %CV Ke %CV Rmax, IIV %CV 9.4 (42.9) 21.8 (54.9) Rmax, IOV %CV 17.6 (44.7) 15.4 (59.3) EC50 ICScort %CV 60.0 (47.1) 46.0 (158.0) Kin %CV n/c n/c Kout %CV 7.8 (162.6) 6.3 (127.9) EC50 Cortlym %CV 35.6 (69.0) 71.7 (80.4) EC50 ICSlym %CV 80.4 (36.6) 88.7 (133.3) cortisol ng/ml 0.54 (8.7) 0.54 (10.0) lymphocytes % 0.17 (12.2) 0.17 (15.2)

PAGE 83

71 Administration Time 8am9am10am11am12am1pm2pm3pm4pm5pm6pm7pm8pm9pm10pm11pm12pm1am2am3am4am5am6am7am Cumulative Cortisol Suppression (%) -1001020304050 Administration Time 8am9am10am11am12am1pm2pm3pm4pm5pm6pm7pm8pm9pm10pm11pm12pm1am2am3am4am5am6am7am Cumulative Cortisol Suppression (%) -1001020304050 ab Administration Time 8am9am10am11am12am1pm2pm3pm4pm5pm6pm7pm8pm9pm10pm11pm12pm1am2am3am4am5am6am7am Cumulative Lmphoctes Suppression (%) 05101520 c Administration Time 8am9am10am11am12am1pm2pm3pm4pm5pm6pm7pm8pm9pm10pm11pm12pm1am2am3am4am5am6am7am 05101520 Cumulative Lmphoctes Suppression (%)d Figure 5-5. Relationship between administration time and corresponding mean (+ SE) cumulative suppression. a) CCS of the 75 subjects after once-daily 1mg of BUD. b) CCS of the 75 subjects after once-daily 2mg of BUD. c) CLS of the 75 subjects after once-daily 1mg of BUD. d) CLS of the 75 subjects after once-daily 2mg of BUD. Empty circle: simulation 1; Filled circle: simulation 2; Empty triangle: simulation 3; Filled triangle: simulation 4; Empty Square: simulation 5; Filled Square: simulation 6; Empty heart: simulation 7; Filled heart: simulation 8.

PAGE 84

72 Administration Time 8am9am10am11am12am1pm2pm3pm4pm5pm6pm7pm8pm9pm10pm11pm12pm1am2am3am4am5am6am Cumulative Cortisol Suppression (%) -1001020304050 Administration Time 8am9am10am11am12am1pm2pm3pm4pm5pm6pm7pm8pm9pm10pm11pm12pm1am2am3am4am5am6am7am Cumulative Cortisol Suppression (%) -1001020304050 ab Administration Time 8am9am10am11am12am1pm2pm3pm4pm5pm6pm7pm8pm9pm10pm11pm12pm1am2am3am4am5am6am7am Cumulative Lmphoctes Suppression (%) 05101520 c Administration Time 8am9am10am11am12am1pm2pm3pm4pm5pm6pm7pm8pm9pm10pm11pm12pm1am2am3am4am5am6am7am 05101520 Cumulative Lmphoctes Suppression (%)d Figure 5-6. Relationship between administration time and corresponding mean (+ SE) cumulative suppression. a) CCS of the 75 subjects after once-daily 0.5mg of FP. b) CCS of the 75 subjects after once-daily 1mg of FP. c) CLS of the 75 subjects after once-daily 0.5mg of FP. d) CLS of the 75 subjects after once-daily 1mg of FP. Empty circle: simulation 1; Filled circle: simulation 2; Empty triangle: simulation 3; Filled triangle: simulation 4; Empty Square: simulation 5; Filled Square: simulation 6; Empty heart: simulation 7; Filled heart: simulation 8.

PAGE 85

73 Table 5-4. Summary of the mean cumulative cortisol suppression of the 75 subjects of the 8 simulations Corticosteroid Maximum SD a (%) Administration time point b Minimum SD c (%) Administration time point c Fluctuation d (%) 1 mg 27.5 0.9 2am 5am -3.5 1.9 3pm 6pm 31.0 2.5 BUD 2 mg 44.0 1.6 3am 5am 12.5 1.2 4pm 5pm 31.5 2.0 0.5 mg 12.9 2.6 4am 8am -4.4 2.4 5pm 8pm 17.3 3.3 FP 1 mg 22.1 1.3 4am 5am -0.2 1.4 3pm 6pm 22.3 3.1 a. The mean and standard deviation of the maximum mean CCS of the 75 subjects of the 8 simulations. b. The range of administration time points which corresponds to the maximum mean CCS of the 75 subjects of the 8 simulations. c. The mean and standard deviation of the minimum mean CCS of the 75 subjects of the 8 simulations. d. The range of administration time points which corresponds to the minimum mean CCS of the 75 subjects of the 8 simulations. e. The mean and standard deviation of the difference between maximum mean CCS and the minimum mean CCS of the 75 subjects of the 8 simulations. Table 5-5. Summary of the mean cumulative lymphocytes suppression of the 75 subjects of the 8 simulations Corticosteroid Maximum SD a (%) Administration time point b Minimum SD c (%) Administration time point c Fluctuation d (%) 1 mg 9.6 0.5 4pm 8pm 4.7 0.6 5am 9am 4.9 0.6 BUD 2 mg 15.8 0.7 3pm 8pm 9.8 1.0 3am 5am 6.0 1.5 0.5 mg 6.8 0.3 4pm 7pm 2.5 0.4 3am 8am 4.2 0.5 FP 1 mg 9.6 0.4 3pm 6pm 4.5 0.4 5am 9am 5.2 0.4 a. The mean and standard deviation of the maximum mean CLS of the 75 subjects of the 8 simulations. b. The range of administration time points which corresponds to the maximum mean CLS of the 75 subjects of the 8 simulations. c. The mean and standard deviation of the minimum mean CLS of the 75 subjects of the 8 simulations. d. The range of administration time points which corresponds to the minimum mean CLS of the 75 subjects of the 8 simulations. e. The mean and standard deviation of the difference between maximum mean CLS and the minimum mean CLS of the 75 subjects of the 8 simulations. Discussion Without exogenous stimuli, both endogenous cortisol concentrations and lymphocyte counts undergo a circadian rhythm. Moreover, the circadian rhythm of lymphocyte counts is inversely correlated with the circadian rhythm of endogenous

PAGE 86

74 cortisol [121, 144]. One therefore should expect that administration time of an ICS has an effect on not only the cortisol but also lymphocytes suppression. In the current study we employed a population PK/PD simulation approach to evaluate the effect of the administration time on cumulative cortisol suppression and cumulative lymphocytes suppression simultaneously, and to identify an optimum administration time point for once-a-day dosing regimen in terms of minimum CCS and CLS, respectively. A population PK/PD model was developed in this study to describe the disposition of BUD and FP after inhalation, the effect of BUD and FP on cortisol, and the combined effect of BUD or FP and cortisol on lymphocytes in healthy subjects. The structure (PK/PD) model was a mechanism-based model and previous study has demonstrated the validity of the model via standard two-stage analysis [121]. The current study successfully applied the model to non-linear mixed effects modeling analysis and estimated the inter and intra-subject and inter-occasion variability of model. After repeated inhalation, BUD plasma concentrations time-profiles were well described by one-compartment model with first-order absorption, while FP plasma concentrations time-profile was well described by two-compartment model with first-order absorption. The extra disposition of FP to a peripheral compartment compared with BUD could result from the more sensitive assay, about 5 folds more sensitive than assay for BUD, and its higher lipophilicity than BUDs. The PK parameters of BUD and FP derived from this study are in broad agreement with previously reported values [145]. The steroid-independent pharmacodynamic parameters (Tmax, Tmin, Rmax, Kin and Kout), EC50ICScort of BUD and FP, EC50ICSlym of BUD and EC50Cortlym observed with the non-linear mixed effect modeling approach are comparable to results of previous

PAGE 87

75 studies [75, 121, 146]. A linear relationship between EC50 for cortisol suppression for various corticosteroids and their relative receptor affinity (RRA), estimated in vitro, has been established in previous studies [147]. Hence, a linear relationship between EC50 for lymphocytes suppression for various corticosteroids and their RRA is reasonably expected. If we plot the 1/RRA versus EC50 of BUD, cortisol and FP for suppression on lymphocytes obtained in this study, a good linear curve (R2 = 0.95) could be established and should prove useful in future prediction of EC50 of other steroids (in vivo effects) based on their in vitro RRA (Figure 5-7). 503005508001050130015501800RRA (Relative Receptor Affinity) -23813181/EC50free,lym (ng/ml) FPCortisolBUD1/EC50 = -1.373 + 0.01109*RRA Figure 5-7. Relationship between EC 50, free of steroids for lymphocytes suppression and their reciprocal of the relative receptor affinity.

PAGE 88

76 For both drugs, the maximum CCS was observed when ICS was administered in the early morning and the minimum CCS was observed when ICS was administered in the afternoon, and this observation is in agreement with previous report [127]. This circadian rhythm results from the interaction between the systemic activity of exogenous corticosteroids and endogenous cortisol release: CCS is maximized when the period of high systemic activity of exogenous corticosteroids (concentrations higher than EC50) covers the period of high cortisol release in the early morning; CCS is minimized when the period of high systemic activity covers the period of low cortisol release in the late night. On the contrary, the maximum CLS was observed when ICS was given in the afternoon while the minimum CLS was observed when ICS was given in the early morning. This circadian rhythm is the result of combined modulation of natural circadian rhythm of the lymphocytes release by systemic activity of exogenous corticosteroids and endogenous cortisol, whose release is suppressed by the same systemic exogenous corticosteroids. Hence, the CLS reaches the maximum when the period of high systemic activity of the net corticosteroids, the total of the suppressed endogenous cortisol and exogenous corticosteroids, covers the period of high lymphocytes release in the late night. And it reaches the minimum when the period of high systemic activity of the net corticosteroids covers the period of low lymphocytes release in the early morning. So for once-a-day dosing regimen, the optimum administration time is in the afternoon in terms of minimized CCS, however, the optimum administration time is in the morning in terms of minimized CLS. In our simulations, the administration time for which maximum suppression of cortisol and blood lymphocytes was observed differed. The advantage to use cortisol

PAGE 89

77 suppression as biomarker is that cortisol is more sensitive to the presence of exogenous corticosteroids since the EC50 values for the effect of ICS on lymphocytes was about 1 -8 times higher than that for the effect on cortisol as demonstrated in previous and this study [146, 148]. It is, however, disputed whether cortisol suppression is a good marker for systemic effects or side effects, as the mechanism of cortisol suppression is likely to not reflect anti-inflammatory events or effects on growth that is affected by both endogenous and exogenous glucocorticoids. While the lymphocytes suppression is somewhat less sensitive than the serum cortisol suppression, the immunological character of lymphocytopenia, has the advantage of representing a more relevant pharmacodynamic endpoint, that is affected by both endogenous and exogenous glucocorticoids. Lymphocytopenia is a systemic effect, and should represent a side effect for inhalation therapy. According to this definition delivery of ICS in the evening hours would induce the highest degree of side effects. Similar, administration of a systemic glucocorticoid (e.g. oral delivery) would induce the maximum effects when given in the evening. On the other hand, it is also likely that the desired pulmonary effects of inhaled glucocorticoids are the sum of endogenous and exogenous glucocorticoids. Consequently, one would also predict maximum antiasthmatic effects for an evening administration. The differences observed in the simulations between the two biomarkers supports the note that these biomarkers are not interchangeable on a qualitative and quantitative level. The terminal half-life of FP after inhalation is much longer than that of BUD (14.3 hours vs. 3.3 hours as estimated from this study). It has been recognized that for a receptor mediated response, the duration and decline of this response depends partly on

PAGE 90

78 the elimination rate of the drug [149, 150]. For FP, a longer terminal life leads to less fluctuated concentration during the dosing interval fluctuate less. This resulted in our dosing time simulations in a smaller difference between maximum and minimum cumulative cortisol suppression for 500 and 1000 g FP than for 1,000 and 2,000 g BUD (Table 5-4). Thus, the dosing time is less important for the CCS of FP. Interestingly, the difference between maximum and minimum CLS after once-daily administration of 1mg or 2mg BUD and after once-daily of 0.5 or 1mg FP were rather similar (Table 5-4), which is likely be related to the rather small overall effect on blood lymphocytes and the buffering capacity of cortisol. Once-a-day dosing has been successfully evaluated for some of the newer inhaled glucocorticoids (budesonide, ciclesonide) because of the easy compliance [48]. Our results indicate that the administration time may be important for once-a-day dosing regimen and that a late afternoon or evening administration is likely to result in maximum effects. Clinical studies evaluating the effect of time of dosing on ICS efficacy and side effects are limited. Some clinical studies suggested that afternoon or early evening dosing tends to be superior in efficacy to a morning dosing especially for patients with nocturnal asthma [151, 152]. This is in agreement with our simulation results when data on systemic lymphocytopenia are applied to the local pulmonary effects. Conclusion The population PK/PD modeling/simulation presented in current study is a step further into predicting outcome of clinical study based on previous developed PK/PD model. The population approach taken in our study incorporates both inter-subject and intra-subject variability, and allows probing for significant differences. Previous non-population PK/PD simulation approaches have already proven high predictive power for

PAGE 91

79 CCS [153], our population PK/PD simulation should be even more informative with respect to the statistical evaluation of the simulated events. Employing this approach, we demonstrated the administration time is an important factor to minimize the systemic side effects and maximize the benefit/risk ratio for once-daily regimen of ICS. However, further clinical studies are needed to verify the prediction by our simulation.

PAGE 92

CHAPTER 6 CONCLUSIONS Pulmonary targeting is a desired characteristic for successful inhaled corticosteroids. The overall goal of this study was to evaluate certain aspects of factors affecting pulmonary targeting of current ICSs, hence to gain more understanding of those factors and contribute to the journey of developing ideal ICS for asthmatic patients. Our first study investigated how different valved holding chambers (VHCs) affect fluticasone propionate aerosols delivered from a metered dose inhaler (MDI). It was found that all three VHCs tested, AeroChamber-Plus OptiChamber and OptiChamber -Advantage significantly reduced the in vitro oropharyngeal deposition of fluticasone propionate. It was also found that the respirable dose from the MDI attached to AeroChamber-Plus or OptiChamber was not significantly different from the respirable dose from fluticasone propionate MDI alone, while the respirable dose of fluticasone from OptiChamber -Advantage was 40-45% less than from the MDI alone. This result suggests that AeroChamber-Plus or OptiChamber could be successfully used with fluticasone MDI since both VHCs not only could reduce the risk of topical adverse effects induced by fluticasone, but also deliver the same amount of respirable dose of fluticasone, while OptiChamber -Advantage would not be a good choice for patients on fluticasone MDI since much less of respirable dose of fluticasone was delivered by it. This study shows that choice of VHC is an important factor for determination of pulmonary targeting of fluticasone propionate. 80

PAGE 93

81 Our second study showed that the systemic absorption of beclometasone monopropionate, budesonide, fluticasone propionate and mometasone furoate appears to be greater in patients with a higher FEV 1 % predicted although other factors, including age and gender affect this relationship. In the third study, using a methacholine challenge we were able to study the effect of change in lung function within an individual on plasma concentrations of fluticasone and budesonide under controlled conditions. A 33 % reduction in FEV 1 in our study caused a 60 % reduction in AUC 0-5h values for fluticasone and 29% reduction in AUC 0-5 values for budesonide, suggesting a 60 % reduction and a 29% in systemic exposure for fluticasone and budesonide respectively. Results from these two studies indicate that the systemic exposure, hence the risk adverse systemic effects from ICSs is correlated with asthmatic patients lung function, and extent of the correlation varies for different ICSs. We suggest that a difference in lipophilicity among ICSs probably is the reason for the difference in the extent of correlation among various ICSs. Further study is necessary to explore this. Our findings re-enforce the importance of reviewing the need for higher doses of inhaled corticosteroids and particularly fluticasone, as lung function improves and in patients with relatively normal lung function, to improve the benefit to risk ratio of ICSs. Use of biomarker, such as cortisol suppression and lymphocytes facilitates the evaluation of systemic effects of different ICS regimens. Population PK/PD modeling and simulation provides us a great tool to describe the relationship between drug concentrations and biomarker concentrations, and to predict and assess statistical significance of the outcomes of future studies. Hence, we developed a population PK/PD model for inhaled budesonide and fluticasone, respectively to describe the relationship

PAGE 94

82 between drug concentrations, cortisol concentrations and lymphocyte counts. Further, we used this model to evaluate the administration time effect on the cumulative cortisol suppression (CCS) and lymphocytes suppression (CLS) for once-daily inhaled BUD and FP, respectively. Our results indicates that the optimal time for administration of BUD or FP, in terms of minimized systemic side effects, depends on the choice of the biomarker: the minimum CCS is achieved when ICS is given in the afternoon, while minimum CLS is achieved when ICS is given in the morning. This administration time effect is the result of interaction between natural circadian rhythm of endogenous cortisol concentration and lymphocyte concentrations, and the effect of PK/PD properties of ICSs on these two rhythms. Our study also showed that both the difference between maximum CCS and minimum CCS, and the difference between maximum CLS and minimum CLS were significant (P <0.001). The extent of difference between maximum and minimum CCS suppression differs between BUD and FP because their different PK properties. However, whether the difference between maximum and minimum suppression is clinically significant needs further clinical studies to evaluate.

PAGE 95

APPENDIX A NONMEM CODE FOR PK MODEL OF BUD AFTER INHALATION 83

PAGE 96

APPENDIX B NONMEM CODE FOR PK MODEL OF FP AFTER INHALATION 84

PAGE 97

APPENDIX C NONMEM CODE FOR PK/PD MODEL OF BUD 85

PAGE 98

86

PAGE 99

87

PAGE 100

88

PAGE 101

89

PAGE 102

90

PAGE 103

91

PAGE 104

APPENDIX D NONMEM CODE FOR PK/PD MODEL OF FP 92

PAGE 105

93

PAGE 106

94

PAGE 107

95

PAGE 108

96

PAGE 109

97

PAGE 110

APPENDIX E NONMEM CODE FOR BUD PK/PD SIMULATION MODEL 98

PAGE 111

99

PAGE 112

100

PAGE 113

101

PAGE 114

102

PAGE 115

APPENDIX F NONMEM CODE FOR FP PK/PD SIMULATION MODEL 103

PAGE 116

104

PAGE 117

105

PAGE 118

106

PAGE 119

107

PAGE 120

LIST OF REFERENCES 1. National Asthma Education and Prevention Program. Expert Panel Report 2. Guidelines for diagnosis and management of asthma. US Department of Health and Human Services, National Institutes of Health: Bethesda, MD. 1997. NIH Publication no 97-405: p. 288-294. 2. Hartert TV, P.R.J., Epidemiology of asthma: the year in review. Curr. Opin. Pulm. Med., 2000. 6(1): p. 4-9. 3. Barnes, P.J., Pedersen, S., Busse, W.W., Efficacy and safety of inhaled corticosteroids: new developments. Am. J. Respi. Crit. Care Med., 1998. 157(3): p. S1-S53. 4. Wilcox, J.B. and G.S. Avery, Beclomethasone dipropionate corticosteroid inhaler: a preliminary report of its pharmacological properties and therapeutic efficacy in asthma. Drugs, 1973. 6(2): p. 84-93. 5. Szefler, S.J., Pharmacokinetics of intranasal corticosteroids. J. Allergy Clin. Immunol., 2001. 108(Suppl. 1): p. S26-31. 6. Dietzel, K., R. Engelstatter, and A. Keller, Ciclesonide: An on-site activated steroid. Prog. Respir. Res., 2001. 31(10): p. 91-93. 7. Lemanske, R.F., Jr., Busse, W. W., Asthma. J. Allergy Clin. Immunol., 2003. 111( Suppl 2): p. S502-19. 8. Sedgwick, J.B., Calhoun, W. J., Gleich, G. J., Kita, H., Abrams, J. S., Schwartz, L. B., Volovitz, B., Ben-Yaakov, M., Busse, W. W., Immediate and late airway response of allergic rhinitis patients to segmental antigen challenge. Characterization of eosinophil and mast cell mediators. Am. Rev. Respir. Dis., 1991. 144(6): p. 1274-81. 9. Bradding, P., Okayama, Y., Howarth, P. H., Church, M. K., Holgate, S. T., Heterogeneity of human mast cells based on cytokine content. J. Immunol., 1995. 155(1): p. 297-307. 108

PAGE 121

109 10. Shaver, J.R., Zangrilli, J. G.,Cho, S. K.,Cirelli, R. A.Pollice, M.,Hastie, A. T.,Fish, J. E.,Peters, S. P., Kinetics of the development and recovery of the lung from IgE-mediated inflammation: dissociation of pulmonary eosinophilia, lung injury, and eosinophil-active cytokines. Am. J. Respir. Crit. Care Med., 1997. 155(2): p. 442-8. 11. Bousquet, J., Chanez, P., Lacoste, J. Y., Barneon, G., Ghavanian, N., Enander, I., Venge, P., Ahlstedt, S., Simony-Lafontaine, J., Godard, P., Eosinophilic inflammation in asthma. N. Engl. J. Med., 1990. 323(15): p. 1033-9. 12. Barnes, P.J., Pedersen, S., Efficacy and safety of inhaled corticosteroids in asthma. Report of a workshop held in Eze, France, October 1992 Am. Rev. Respir. Dis., 1993. 148(4): p. S1-26. 13. Barnes, P.J., Inhaled glucocorticoids for asthma. N. Engl. J. Med., 1995. 332(13): p. 868-75. 14. Rabe, K.F. and D.T. Schmidt, Pharmacological treatment of asthma today. Eur. Respir. J. 2001. 34(Suppl.): p. 34-40. 15. Reed, C.E., New therapeutic approaches in asthma. J. Allergy Clin. Immunol., 1986. 77(4): p. 537-43. 16. Pelaia, G., Vatrella, A., Calabrese, C., Mazzarella, G., Marsico, S. A., New perspectives in asthma treatment. Allergy, 2000. 55 (Suppl. 61): p. 60-6. 17. Drazen, J.M., E. Israel, and P.M. O'Byrne, Treatment of asthma with drugs modifying the leukotriene pathway. N. Engl. J. Med., 1999. 340(3): p. 197-206. 18. Horwitz, R.J., K.A. McGill, and W.W. Busse, The role of leukotriene modifiers in the treatment of asthma. Am. J. Respir. Crit. Care Med., 1998. 157(5): p. 1363-71. 19. Umland, S.P., R.P. Schleimer, and S.L. Johnston, Review of the molecular and cellular mechanisms of action of glucocorticoids for use in asthma. Pulm. Pharmacol. Ther., 2002. 15(1): p. 35-50. 20. Oakley, R.H., Jewell, C. M., Yudt, M. R., Bofetiado, D. M., Cidlowski, J. A., The dominant negative activity of the human glucocorticoid receptor beta isoform. Specificity and mechanisms of action. J. Biol. Chem., 1999. 274(39): p. 27857-66. 21. Smith, K.J., H.K. Chan, and K.F. Brown, Influence of flow rate on aerosol particle size distributions from pressurized and breath-actuated inhalers. J. Aerosol Med., 1998. 11(4): p. 231-45. 22. Bertorelli, G., V. Bocchino, and D. Olivieri, Heat shock protein interactions with the glucocorticoid receptor. Pulm. Pharmacol. Ther., 1998. 11(1): p. 7-12.

PAGE 122

110 23. Mitchell, P.J. and R. Tjian, Transcriptional regulation in mammalian cells by sequence-specific DNA binding proteins. Science, 1989. 245(4916): p. 371-8. 24. Flower, R.J. and N.J. Rothwell, Lipocortin-1: cellular mechanisms and clinical relevance. Trends Pharmacol. Sci., 1994. 15(3): p. 71-6. 25. Keatings, V.M., Jatakanon, A., Worsdell, Y. M., Barnes, P. J., Effects of inhaled and oral glucocorticoids on inflammatory indices in asthma and COPD. Am. J. Respir. Crit. Care Med., 1997. 155(2): p. 542-8. 26. Barnes, P.J., Molecular mechanisms and cellular effects of glucocorticosteroids. Immunol. Allergy Clin. North. Am., 2005. 25(3): p. 451-68. 27. Mackie, A.E., McDowall, J. E., Ventresca, P., Bye, A., Falcoz, C., Daley-Yates, P. T., Systemic exposure to fluticasone propionate administered via metered-dose inhaler containing chlorofluorocarbon or hydrofluoroalkane propellant. Clin. Pharmacokinet., 2000. 39 (Suppl 1): p. 17-22. 28. Shaw, R.J., Inhaled corticosteroids for adult asthma: impact of formulation and delivery device on relative pharmacokinetics, efficacy and safety. Respir. Med., 1999. 93(3): p. 149-60. 29. Borgstrom, L., The pharmacokinetics of inhaled hydrofluoroalkane formulations. J. Allergy Clin. Immunol., 1999. 104(6): p. S246-9. 30. Leach, C.L., P.J. Davidson, and R.J. Boudreau, Improved airway targeting with the CFC-free HFA-beclomethasone metered-dose inhaler compared with CFC-beclomethasone. Eur. Respir. J., 1998. 12(6): p. 1346-53. 31. Weinbrenner, A., Huneke, D.,Zschiesche, M.,Engel, G.,Timmer, W.,Steinijans, V. W.,Bethke, T.,Wurst, W.,Drollmann, A.,Kaatz, H. J.,Siegmund, Circadian rhythm of serum cortisol after repeated inhalation of the new topical steroid ciclesonide. J. Clin. Endocrinol. Metab., 2002. 87(5): p. 2160-3. 32. Nolting, A., S. Sista, and W. Abramowitz, Single-dose study to compare the pharmacokinetics of HFA flunisolide and CFC flunisolide. J. Pharm. Sci., 2002. 91(2): p. 424-32. 33. Kelly, H.W., Establishing a therapeutic index for the inhaled corticosteroids: part I. Pharmacokinetic/pharmacodynamic comparison of the inhaled corticosteroids. J. Allergy Clin. Immunol., 1998. 102(4): p. S36-51. 34. Derendorf H., H.G., Krishnaswami S., Mollman H.W., Systemic disposition and effects of inhaled corticosteroids. Lung Biology in Health and Disease ed. O.B.P. Schleimer RP, Szefler SJ,Brattsand R. Vol. Inhaled Steroids in Asthma. 2002, New York: Marcel Dekker, Inc. 216-219.

PAGE 123

111 35. Pauwels, R., S. Newman, and L. Borgstrom, Airway deposition and airway effects of antiasthma drugs delivered from metered-dose inhalers. Eur. Respir. J., 1997. 10(9): p. 2127-38. 36. Borgstrom, L., Bondesson, E., Moren, F., Trofast, E., Newman, S. P., Lung deposition of budesonide inhaled via Turbuhaler: a comparison with terbutaline sulphate in normal subjects. Eur. Respir. J., 1994. 7(1): p. 69-73. 37. Placke, M.E., Zimlich, W.C. Jr., Developing an advanced pulmonary delivery technology: leveraging engineering challenges into clinical opportunities. Drug Delivery Technologies, 2002. 2(2): p. 59-63.. 38. Newman, S.P., Brown, J., Steed, K. P., Reader, S. J., Kladders, H., Lung deposition of fenoterol and flunisolide delivered using a novel device for inhaled medicines: comparison of RESPIMAT with conventional metered-dose inhalers with and without spacer devices. Chest, 1998. 113(4): p. 957-63. 39. Nave, R., K. Zech, and T.D. Bethke, Lower oropharyngeal deposition of inhaled ciclesonide via hydrofluoroalkane metered-dose inhaler compared with budesonide via chlorofluorocarbon metered-dose inhaler in healthy subjects. Eur. J. Clin. Pharmacol., 2005. 61(3): p. 203-8. 40. Bender, B., Milgrom, H., Compliance with asthma therapy: a case for shared responsibility. J. Asthma, 1996. 33(4): p. 199-202. 41. Reinberg, A., Halberg, F., Falliers, C. J. Circadian timing of methylprednisolone effects in asthmatic boys. Chronobiologia, 1974. 1(4): p. 333-47. 42. Reinberg, A., Gervais, P., Chaussade, M., Fraboulet, G., Duburque, B., Circadian changes in effectiveness of corticosteroids in eight patients with allergic asthma. J. Allergy Clin. Immunol., 1983. 71(4): p. 425-33. 43. Beam, W.R., Weiner, D. E.,Martin, R. J., Timing of prednisone and alterations of airways inflammation in nocturnal asthma. Am. Rev. Respir. Dis., 1992. 146(6): p. 1524-30. 44. Nichols, T., C.A. Nugent, and F.H. Tyler, Diurnal variation in suppression of adrenal function by glucocorticoids. J. Clin. Endocrinol. Metab., 1965. 25(3): p. 343-9. 45. Segre, E.J. and E.L. Klaiber, Therapeutic utilization of the diurnal variation in pituitary-adrenocortical activity. Calif. Med., 1966. 104(5): p. 363-5. 46. Grant, S.D., P.H. Forsham, and V.C. DiRaimondo, Suppression of 17-hydroxycorticosteroids in plasma and urine by single and divided doses of triamcinolone. N. Engl. J. Med., 1965. 273(21): p. 1115-8.

PAGE 124

112 47. Reinberg, A., Guillet, P., Gervais, P., Ghata, J., Vignaud, D., Abulker, C., One month chronocorticotherapy (Dutimelan 8 15 mite). Control of the asthmatic condition without adrenal suppression and circadian rhythm alteration. Chronobiologia, 1977. 4(4): p. 295-312. 48. Boulet, L.P., Once-daily inhaled corticosteroids for the treatment of asthma. Curr. Opin. Pulm. Med., 2004. 10(1): p. 15-21. 49. Lipworth, B.J. and C.M. Jackson, Safety of inhaled and intranasal corticosteroids: lessons for the new millennium. Drug Saf., 2000. 23(1): p. 11-33. 50. Weiner, P., Berar-Yanay, N., Davidovich, A., Magadle, R., Nocturnal cortisol secretion in asthmatic patients after inhalation of fluticasone propionate. Chest, 1999. 116(4): p. 931-4. 51. Brutsche MH, B.I., Munavvar M, Langley SJ, Masterson CM, Daley-Yates PT, Brown R, Custovic A, Woodcock A., Comparison of pharmacokinetics and systemic effects of inhaled fluticasone propionate in patients with asthma and healthy volunteers: A randomised crossover study. Lancet 2000. 356(9229): p. 556-561. 52. Nave, R., Fisher, R., Zech, K., In vitro metabolism of ciclesonide in human lung and liver precision-cut tissue slices. Biopharm. Drug Dispos., 2006. 27(4): p. 197-207. 53. Richter, K., Kanniess, F., Biberger, C., Nave, R., Magnussen, H., Comparison of the oropharyngeal deposition of inhaled ciclesonide and fluticasone propionate in patients with asthma. J. Clin. Pharmacol., 2005. 45(2): p. 146-52. 54. Harrison, T.W., Systemic availability of inhaled budesonide and fluticasone propionate: healthy versus asthmatic lungs. BioDrugs, 2001. 15(6): p. 405-11. 55. Harrison, T.W. and A.E. Tattersfield, Plasma concentrations of fluticasone propionate and budesonide following inhalation from dry powder inhalers by healthy and asthmatic subjects. Thorax, 2003. 58(3): p. 258-60. 56. Burton, J.A., Schanker, L. S., Absorption of corticosteroids from the rat lung. Steroids, 1974. 23(5): p. 617-24. 57. Gonzalez-Rothi, R.J., Suarez, S., Hochhaus, G., Schreier, H., Lukyanov, A., Derendorf, H., Costa, T. D., Pulmonary targeting of liposomal triamcinolone acetonide phosphate. Pharm. Res., 1996. 13(11): p. 1699-703. 58. Suarez, S., Gonzalez-Rothi, R. J.,Schreier, H.,Hochhaus, G., Effect of dose and release rate on pulmonary targeting of liposomal triamcinolone acetonide phosphate. Pharm. Res., 1998. 15(3): p. 461-5.

PAGE 125

113 59. Ben-Jebria, A., Chen, D., Eskew, M. L., Vanbever, R., Langer, R., Edwards, D. A., Large porous particles for sustained protection from carbachol-induced bronchoconstriction in guinea pigs. Pharm. Res., 1999. 16(4): p. 555-61. 60. Edwards, D.A., A. Ben-Jebria, and R. Langer, Recent advances in pulmonary drug delivery using large, porous inhaled particles. J. Appl. Physiol., 1998. 85(2): p. 379-85. 61. Nave, R., Meyer, W., Fuhst, R., Zech, K., Formation of fatty acid conjugates of ciclesonide active metabolite in the rat lung after 4-week inhalation of ciclesonide. Pulm. Pharmacol. Ther., 2005. 18(6): p. 390-6. 62. Edsbacker, S. and R. Brattsand, Budesonide fatty-acid esterification: a novel mechanism prolonging binding to airway tissue. Review of available data. Ann. Allergy Asthma Immunol., 2002. 88(6): p. 609-16. 63. Derendorf, H., G. Hochhaus, and H. Mollmann, Evaluation of pulmonary absorption using pharmacokinetic methods. J. Aerosol Med., 2001. 14 (Suppl. 1): p. S9-17. 64. C. Falcoz, A.E.M., J. McDowall, J. McRae, L. Yogendran, G.P. Ventresca, A. Bye, Oral bioavailability of fluticasone propionate in healthy subjects. Br. J. Clin. Pharmacol., 1996. 41(23): p. 459P-60P. 65. Daley-Yates, P.T., Price, A.C., Sisson, J.R., Beclomethasone dipropionate: absolute bioavailability, pharmacokinetics and metabolism following intravenous, oral, intranasal and inhaled administration in man. Br. J. Clin. Pharmacol., 2001. 51(5): p. 400-9. 66. Onrust, S.V. and H.M. Lamb, Mometasone furoate. A review of its intranasal use in allergic rhinitis. Drugs, 1998. 56(4): p. 725-45. 67. Ryrfeldt, A., Anderson, P., Edsbacker, S., Tonnesson, M., Pharmacokinetics and metabolism of budesonide, a selective glucocorticoid. Eur. J. Respir. Dis., 1982. 122(Suppl): p. 86-95. 68. Chaplin, M.D., Rooks, W. d, Swenson, E. W., Cooper, W. C., Nerenberg, C., Chu, N. I., Flunisolide metabolism and dynamics of a metabolite. Clin. Pharmacol. Ther., 1980. 27(3): p. 402-13. 69. Derendorf, H., Hochhaus, G., Rohatagi, S., Mollmann, H.,Barth, J.,Sourgens, H.,Erdmann, M., Pharmacokinetics of triamcinolone acetonide after intravenous, oral, and inhaled administration. J. Clin. Pharmacol., 1995. 35(3): p. 302-5.

PAGE 126

114 70. Nave, R., Bethke, T., Jvan Marle, S.P., Zech, K., Pharmacokinetics of 14C-Ciclesonide after Oral and Intravenous Administration in Healthy Subjects. Clin. Pharmacokinet., 2004. 43(7): p. 479-86. 71. Cochrane, M.G., Bala, M. V., Downs, K. E., Mauskopf, J., Ben-Joseph, R. H., Inhaled corticosteroids for asthma therapy: patient compliance, devices, and inhalation technique. Chest, 2000. 117(2): p. 542-50. 72. Crim, C., L.N. Pierre, and P.T. Daley-Yates, A review of the pharmacology and pharmacokinetics of inhaled fluticasone propionate and mometasone furoate. Clin. Ther., 2001. 23(9): p. 1339-54. 73. Ryrfeldt, A., Anderson, P.,Edsbacker, S., Tonnesson, M., Pharmacokinetics and metabolism of budesonide, a selective glucocorticoid. Eur. J. Respir. Dis., 1982. 122(Suppl.): p. 86-95. 74. Mackie, A.E., Ventresca, G. P., Fuller, R. W., Bye, A., Pharmacokinetics of intravenous fluticasone propionate in healthy subjects. Br. J. Clin. Pharmacol., 1996. 41(6): p. 539-42. 75. Rohatagi, S., Arya, V.,Zech, K.,Nave, R.,Hochhaus, G.,Jensen, B. K.,Barrett, J. S., Population pharmacokinetics and pharmacodynamics of ciclesonide. J. Clin. Pharmacol., 2003. 43(4): p. 365-78. 76. Hochhaus, G., Mollmann, H.,Derendorf, H., Gonzalez-Rothi, R. J., Pharmacokinetic/pharmacodynamic aspects of aerosol therapy using glucocorticoids as a model. J. Clin. Pharmacol., 1997. 37(10): p. 881-92. 77. Hubner, M., G. Hochhaus, and H. Derendorf, Comparative pharmacology, bioavailability, pharmacokinetics, and pharmacodynamics of inhaled glucocorticosteroids. Immunol. Allergy Clin. North. Am., 2005. 25(3): p. 469-88. 78. Postma, D.S., Sevette, C.,Martinat, Y.,Schlosser, N.,Aumann, J.,Kafe, H., Treatment of asthma by the inhaled corticosteroid ciclesonide given either in the morning or evening. Eur. Respir. J., 2001. 17(6): p. 1083-8. 79. Rohatagi, S., Luo, Y., Shen, L., Guo, Z., Schemm, C., Huang, Y., Chen, K., King, P., High plasma protein binding of desisobutyryl-ciclesonide, the active metabolite of ciclesonide, A Novel Inhaled Corticosteroid. The AAPS Journal, 2004. 6(4): p. 13-5. 80. Ryrfeldt A, A.P., Edsbaecker S, Tonnesson M, Davies D, PauwelsR, Pharmacoketics and metabolism of budesonide, a selective glucocorticoid. Eur. Respir. J., 1983. 63(Suppl. 122): p. 86-95.

PAGE 127

115 81. Winkler, J., G. Hochhaus, and H. Derendorf, How the lung handles drugs: pharmacokinetics and pharmacodynamics of inhaled corticosteroids. Proc. Am. Thorac. Soc., 2004. 1(4): p. 356-63. 82. Sharpe, M. and B. Jarvis, Inhaled mometasone furoate: a review of its use in adults and adolescents with persistent asthma. Drugs, 2001. 61(9): p. 1325-50. 83. Ahrens R., L.C., Bahl T., Han S., Choosing the metered-dose inhaler spacer or holding chamber hat matches the patient's need: evidence that specific drug being delivered is an important consideration. J. Allergy Clin. Immunol., 1995. 96(3): p. 288-294. 84. Gunawardena KA, S.T., Jones JI., The spaceholder for delivery of salbutamol: a comparision with the standard metered-dose inhaler plus vomumatic space device. Respir. Med., 1997. 91(21): p. 311-316. 85. Santati, P.A., Clinical comparison of a homemade fruit juice spacer and two standard spacers in children. Am. J. Respir. Crit. Care Med., 2000. 161(2): p. 343-6. 86. Olson, B., Can in vitro measurements predict the lung deposition of dry-powder inhalers? J. Aerosol. Med., 1997. 10(Suppl.): p. 13-16. 87. Pharmacopeia, R., MD., Physical tests powder inhalers. 2001, T. U. S.. 88. Phelan, R.F., Cuddihy, R.G, Fisher G.L., Main features of the prposed NCRP respiratory tract model. Radiat Probt Dosimetry, 1991. 38(4): p. 179-84. 89. Bates, D.V., Fish, B. R., Hatch, T. F., Mercer, T. T., Morrow, P. E., Deposition and retention models for internal dosimetry of the human respiratory tract. Task group on lung dynamics. Health Phys., 1966. 12(2): p. 173-207. 90. Zanen, P., L.T. Go, and J.W. Lammers, Optimal particle size for beta 2 agonist and anticholinergic aerosols in patients with severe airflow obstruction. Thorax, 1996. 51(10): p. 977-80. 91. Bisgaard, H., What dose fraction represents the respirable dose? Respir. Med., 1997. 91(suppl. A): p. 20-1. 92. Pierart, F., Wildhaber, J. H., Vrancken, I., Devadason, S. G., Le Souef, P. N., Washing plastic spacers in household detergent reduces electrostatic charge and greatly improves delivery. Eur. Respir. J., 1999. 13(3): p. 673-8.

PAGE 128

116 93. Asmus, M.J., Coowanitwong, I., Kwon, S. H., Khorsand, N., Hochhaus, G., In vitro performance of two common valved holding chambers with a chlorofluorocarbon-free beclomethasone metered-dose inhaler. Pharmacotherapy, 2003. 23(12): p. 1538-44. 94. Asmus, M.J., Liang, J., Coowanitwong, I., Vafadari, R., Hochhaus, G., In vitro deposition of fluticasone aerosol from a metered-dose inhaler with and without two common valved holding chambers. Ann. Allergy Asthma Immunol., 2002. 88(2): p. 204-8. 95. Davies N., F.M., Brown K., Gipps E., In-vitro characterization of metered dose inhaler versus dry powder inhaler glucocorticoid products: influence of inspiratory flow rates. J. Pharm. Sci., 2000. 3(13): p. 318-324. 96. Brutsche, M.H., Brutsche IC, Munavvar M, Langley SJ, Masterson CM, Daley-Yates PT, Brown R, Custovic A, Woodcock A., Comparison of pharmacokinetics and systemic effects of inhaled fluticasone propionate in patients with asthma and healthy volunteers: A randomised crossover study. Lancet 2000. 356(9229): p. 556-561. 97. Singh, S.D., Whale, C.,Houghton, N.,Daley-Yates, P.,Kirby, S. M.,Woodcock, A. A., Pharmacokinetics and systemic effects of inhaled fluticasone propionate in chronic obstructive pulmonary disease. Br. J. Clin. Pharmacol., 2003. 55(4): p. 375-381. 98. Harrison, T.W., Wisniewski, A., Honour, J., Tattersfield, A. E., Comparison of the systemic effects of fluticasone propionate and budesonide given by dry powder inhaler in healthy and asthmatic subjects. Thorax, 2001. 56(3): p. 186-191. 99. Dimova, H., Wang, Y., Pommery, S., Moellmann, H., Hochhaus, G., SPE/RIA vs LC/MS for measurement of low levels of budesonide in plasma. Biomed. Chromatogr., 2003. 17(1): p. 14-20. 100. Krishnaswami, S., Mollmann, H., Derendorf, H., Hochhaus, G., A sensitive LC-MS/MS method for the quantification of fluticasone propionate in human plasma. J. Pharm. Biomed. Anal., 2000. 22(1): p. 123-129. 101. Wang, Y. and G. Hochhaus, Simultaneous quantification of beclomethasone dipropionate and its metabolite, beclomethasone 17-monopropionate in rat and human plasma and different rat tissues by liquid chromatography-positive electrospray ionization tandem mass spectrometry. Analytical Technologies in the Biomedical & Life Sciences, 2004. 805(2):203-210. 102. Sahasranaman, S., T.Y., Biniasz, D., Hochhaus, G., A sensitive liquid chromatography-tandem mass spectrometry method for the quantification of mometasone furoate in human plasma. J. Chromatography, 2005. 819(1): p. 175-9.

PAGE 129

117 103. Mollmann, H., Wagner, M., Krishnaswami, S., Dimova, S., Tang, Y., Hochhaus, G., Single-dose and steady-state pharmacokinetic and pharmacodynamic evaluation of therapeutically clinically equivalent doses of inhaled fluticasone propionate and budesonide, given as Diskus or Turbohaler dry-powder inhalers to healthy subjects. J Clin Pharmacol, 2001. 41(12): p. 1329-38. 104. Thorsson, L., Edsbacker, S., Kallen, A., Pharmacokinetics and systemic activity of fluticasone via Diskus and pMDI, and of budesonide via Turbuhaler. Br. J. Clin. Pharmacol., 2001. 52(5): p. 529-538 105. Johnson, M., Pharmacodynamics and pharmacokinetics of inhaled glucocorticoids. J. Allergy Clin. Immunol., 1996. 97(1): p. 169-176. 106. Winkler J, H.G., Derendorf H, How the lung handles drugs pharmacokinetics and pharmacodynamics of inhaled corticosteroids. Proc. Am. Thorac. Soc., 2004. 1(4): p. 356-363. 107. Daley-Yates, P., Price, A.C., Sisson, J.R., Beclomethasone dipropionate, absolute bioavailability, pharmacokinetics and metabolism following intravenous, oral intranasal and inhaled administration in man. Br J Clin Pharmacol, 2001. 51(5): p. 400-409. 108. Falcoz, C., Oliver, R., McDowall, J.E., Ventrasca, P., Bioavailability of orally administered micronised fluticasone propionate. Clin. Pharmacokinet., 2000. 39(1): p. 9-15. 109. Daley-Yates, P.T., Price, A. C., Sisson, J. R., Pereira, A., Dallow, N., Beclomethasone dipropionate: absolute bioavailability, pharmacokinetics and metabolism following intravenous, oral, intranasal and inhaled administration in man. Br. J. Clin. Pharmacol., 2001. 51(5): p. 400-9. 110. Agertoft, L., Laulund, L. W., Harrison, L. I., Pedersen, S., Influence of particle size on lung deposition and pharmacokinetics of beclomethasone dipropionate in children. Ped. Pulm., 2003. 35(3): p. 192-199. 111. Harrison, L.I., Kurup, S., Wagner, C., Ekholm, B. P., Larson, J. S., Kaiser, H. B., Pharmacokinetics of beclomethasone 17-monopropionate from a beclomethasone dipropionate extrafine aerosol in adults with asthma. Eur. J. Clin. Pharmacol., 2002. 58(3): p. 197-201.

PAGE 130

118 112. Mollmann, H., Wagner, M., Krishnaswami, S., Dimova, H., Tang, Y., Falcoz, C., Daley-Yates, P. T., Krieg, M., Stockmann, R., Barth, J., Lawlor, C., Mollmann, A. C., Derendorf, H., Hochhaus, G., Single-dose and steady-state pharmacokinetic and pharmacodynamic evaluation of therapeutically clinically equivalent doses of inhaled fluticasone propionate and budesonide, given as Diskus or Turbohaler dry-powder inhalers to healthy subjects. J. Clin. Pharmacol., 2001. 41(12): p. 1329-38. 113. Thorsson, L., Edsbacker S.,Kallen, A.,Lofdahl, C. G., Pharmacokinetics and systemic activity of fluticasone via Diskus and pMDI, and of budesonide via Turbuhaler. Br. J. Clin. Pharmacol., 2001. 52(5): p. 529-38. 114. Affrime, M.B., Cuss, F., Padhi, D., Wirth, M., Pai, S., Clement, R. P., Lim, J., Kantesaria, B., Alton, K., Cayen, M. N., Bioavailability and metabolism of mometasone furoate following administration by metered-dose and dry-powder inhalers in healthy human volunteers. J. Clin. Pharmacol., 2000. 40(11): p. 1227-36. 115. Winkler J, H.G., Derendorf H., How the lung handles drugs pharmacokinetics and pharmacodynamics of inhaled corticosteroids. Proc. Am. Thorac. Soc. 2004. 1(4): p. 356-363. 116. Yan K, S.C., Woolcock AJ., Rapid method for measurement of bronchial responsiveness. Thorax, 1983. 38(10): p. 760-765. 117. Saari SM, V.M., Koskinen MO, Turjanmaa VM, Waldrep JC, Nieminen MM. Regional lung deposition and clearance of 99mTc-labeled beclomethasone-DLPC liposomes in mild and severe asthma. Chest, 1998. 113(6): p. 1573-1579. 118. Thorsson, L., S. Edsbacker, and T.B. Conradson, Lung deposition of budesonide from Turbuhaler is twice that from a pressurized metered-dose inhaler P-MDI. Eur. Respir. J., 1994. 7(10): p. 1839-44. 119. Agertoft L., P.S., Lung deposition and systemic availability of fluticasone Diskus and budesonide Turbuhaler in children. Am. J. Respir. Crit. Care Med., 2003. 168(7): p. 779-782. 120. Rohatagi, S., Moellmann, H., Barth, J., Derendorf, H., Pharmacokinetic interaction between endogenous cortisol and exogenous corticosteroids. Pharmazie, 1995. 50(4): p. 610-13. 121. Meibohm, B., Derendorf, H., Mollmann, H., Frohlich, P., Tromm, A., Wagner, M., Homrighausen, S., Krieg, M., Hochhaus, G., Mechanism-based PK/PD model for the lymphocytopenia induced by endogenous and exogenous corticosteroids Int. J. Clin. Pharmacol. Ther., 1999. 37(8): p. 367-76.

PAGE 131

119 122. Derendorf, H., Hochhaus, G., What is the best marker for inhaled corticosteroid safety? Alle. Asthma Proc., 2005. 26(3): p. 89-94. 123. Meibohm, B., Hochhaus, G., Mollmann, H., Barth, J., Wagner, M., Krieg, M., Stockmann, R., Derendorf, H., A pharmacokinetic/pharmacodynamic approach to predict the cumulative cortisol suppression of inhaled corticosteroids. J. Pharmacokinet. Biopharm., 1999. 27(2): p. 127-47. 124. Honour, J.W., Hypothalamic-pituitary-adrenal axis. Respir. Med., 1994. 88 (Suppl A): p. 9-13; 125. Weitzman, E.D., Fukushima, D.,Nogeire, C.,Roffwarg, H.,Gallagher, T. F.,Hellman, L., Twenty-four hour pattern of the episodic secretion of cortisol in normal subjects. J. Clin Endocrinol. Metab., 1971. 33(1): p. 14-22. 126. Sherman, B., C. Wysham, and B. Pfohl, Age-related changes in the circadian rhythm of plasma cortisol in man. J. Clin. Endocrinol. Metab., 1985. 61(3): p. 439-43. 127. Meibohm, B., Hochhaus, G., Rohatagi, S., Mollmann, H., Barth, J., Wagner, M., Krieg, M., Stockmann, R., Derendorf, H., Dependency of cortisol suppression on the administration time of inhaled corticosteroids. J. Clin. Pharmacol., 1997. 37(8): p. 704-10. 128. Fauci, A., Immunosuppressice and anti-Inflammatory effects of glucocorticoids, in Glucocorticoid Hormone Action, J. Baxter and G. Rousseau, Editors. 1979, Springer: Berlin. p. 449-465. 129. Jennings, B.H., K.E. Andersson, and S.A. Johansson, Assessment of the systemic effects of inhaled glucocorticosteroids: the influence of blood sampling technique and frequency on plasma cortisol and leucocytes. Eur. J. Cli. Pharmacology, 1990. 39(2): p. 127-131. 130. Cupps, T.R., Edgar, L. C., Thomas, C. A., Fauci, A. S., Multiple mechanisms of B cell immunoregulation in man after administration of in vivo corticosteroids. J. Immunol., 1984. 132(1): p. 170-5. 131. Dupont, E., Huygen, K., Schandene, L., Vandercruys, M., Palfliet, K., Wybran, J., Influence of in vivo immunosuppressive drugs on production of lymphokines. Transplantation, 1985. 39(2): p. 143-7. 132. Wofsy, D., Regulation of immunity by anti-T-cell antibodies. West J. Med., 1990. 153(3): p. 265-8. 133. Pytsky, V.I., New aspects in mechanisms of action of corticosteroids. J. Investig. Allergol. Clin. Immunol., 1991. 1(6): p. 359-67.

PAGE 132

120 134. Wilckens, T. and R. De Rijk, Glucocorticoids and immune function: unknown dimensions and new frontiers. Immunol. Today, 1997. 18(9): p. 418-24. 135. Dimova, H., Pharmacokinetic/Pharmacodynamic evaluation of inhaled budesonide after single and repeated administration over five days, in Pharmaceutics. 2001, University of Florida: Gainesville. 136. Krishnaswami, S., Pharmacokinetic/Pharmacodynamic modeling of inhaled corticosteroids, in Pharmaceutics. 2001, University of Florida: Gainesville. 137. Karlsson M. O., S.L.B., The importance of modeling interoccasion variability in population pharmacokinetic analyses. J. Pharmacokinet. Biopharm., 1993. 21(6): p. 735-750. 138. Parke J, H.N., Charles BG., A procedure for generating bootstrap samples for the validation of nonlinear mixed-effects population models. Comput. Methods Programs Biomed., 1999. 59(21): p. 19-29. 139. Rohatagi, S., Bye, A.,Falcoz, C.,Mackie, A. E.,Meibohm, B.,Mollmann, H.,Derendorf, H., Dynamic modeling of cortisol reduction after inhaled administration of fluticasone propionate. J. Clin. Pharmacol., 1996. 36(10): p. 938-41. 140. Derendorf H., M.H., Barth J., Mollmann C., Tunn S., Krieg M., Pharmacokinetics and oral bioavailability of hydrocortisone. J. Clin. Pharmacol., 1991. 31(5): p. 473-6. 141. Ariens, E.J., Affinity and intrinsic activity in the theory of competitive inhibition. I. Problems and theory. Arch. Int. Pharmacodyn. Ther., 1954. 99(1): p. 32-49. 142. Ballard, P., Delivery and transport of glucocorticoids to target cells. Glucocorticoid Hormone Action, ed. R.G. Baxter J. 1979, Nashville, Tennessee Springer. 25-48. 143. Ette, E.I., Stability and performance of a population pharmacokinetic model J. Clin. Pharmacol., 1997. 37(6): p. 486-495. 144. Thomson, S.P., L.J. McMahon, and C.A. Nugent, Endogenous cortisol: a regulator of the number of lymphocytes in peripheral blood. Clin. Immunol. Immunopathol., 1980. 17(4): p. 506-14. 145. Minto, C., Li B., Tattam B., Brown K., Seale J.P.,Donnelly R., Pharmacokinetics of epimeric budesonide and fluticasone propionate after repeat dose inhalation intersubject variability in systemic absorption from the lung. Br. J. Clin. Pharmacol., 2000. 50(2): p. 116-124.

PAGE 133

121 146. Mollmann, H., Wagner, M., Meibohm, B., Hochhaus, G., Barth, J., Stockmann, R., Krieg, M., Weisser, H., Falcoz, C., Derendorf, H., Pharmacokinetic and pharmacodynamic evaluation of fluticasone propionate after inhaled administration. Eur. J. Clin. Pharmacol., 1998. 53(6): p. 459-67. 147. Derendorf H, H.G., Meibohm B, Mollmann H, Bath J, Pharmacokinetics and pharmacodynamics of inhaled corticosteroids. J. Allergy. Clin. Immunol., 1998. 101(4): p. S440-446. 148. Mollmann, H., Derendorf, H., Barth, J., Meibohm, B., Wagner, M., Krieg, M., Weisser, H., Knoller, J., Mollmann, A., Hochhaus, G., Pharmacokinetic/pharmacodynamic evaluation of systemic effects of flunisolide after inhalation. J. Clin. Pharmacol., 1997. 37(10): p. 893-903. 149. Levy, G., Relationship between elimination rate of drugs and rate of decline of their pharmacologic effects. J. Pharm. Sci., 1964. 53(13): p. 342-3. 150. Hochhaus, G., Derendorf H., Dose optimization based on pharmacokinetic-pharmacodynamic modeling. Handbook of Pharmacokinetic/Pharmacodynamic Correlation, ed. H.G. Derendorf H. 1995, Boco Raton, FL: CRC. Press. 79-120. 151. Pincus, D.J., T.R. Humeston, and R.J. Martin, Further studies on the chronotherapy of asthma with inhaled steroids: the effect of dosage timing on drug efficacy. J. Allergy Clin. Immunol., 1997. 100(6): p. 771-4. 152. Noonan, M., Karpel, J. P., Bensch, G. W., Ramsdell, J. W., Webb, D. R., Nolop, K. B., Lutsky, B. N., Comparison of once-daily to twice-daily treatment with mometasone furoate dry powder inhaler. Ann. Allergy Asthma Immunol., 2001. 86(1): p. 36-43. 153. Krishnaswami, S., G. Hochhaus, and H. Derendorf, An interactive algorithm for the assessment of cumulative cortisol suppression during inhaled corticosteroid therapy. AAPS PharmSci., 2000. 2(3): p. E22.

PAGE 134

BIOGRAPHICAL SKETCH Kai Wu was born on April 25, 1977, in Fujian, P. R. China. Kai got his bachelors degree in microbiology from Nankai University in July 1999. Kai entered the graduate school in Southern Illinois University in August 1999 and graduated with a masters degree in molecular biology in December 2001. In August 2002, Kai entered the Ph.D. program at the Department of Pharmaceutics, College of Pharmacy, University of Florida, working under the supervision of Dr. Guenther Hochhaus. In August 2003, Kai entered the masters program at the Department of Statistics, College of Liberal Arts and Sciences, working under the supervision of Dr. Rongling Wu. Kai received his Master of Statistics degree in December 2005 and Doctor of Philosophy degree in pharmaceutics in December 2006. 122