Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 256

______

Clinical-Pharmacokinetic Aspects of Prolonged Effect Duration

as Illustrated by β2-Agonists

BY

JOHAN ROSENBORG

ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2001 Dissertation for the Degree of Doctor of Philosophy (Faculty of Pharmacy) Pharmacokinetics and Drug Therapy presented at Uppsala in 2001

ABSTRACT

Rosenborg, J. 2001. Clinical-Pharmacokinetic Aspects of Prolonged Effect Duration as

Illustrated by β2-Agonists. Acta Univ. Ups. Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 256. 41 pp. Uppsala. ISBN 91-554-5099-7.

Regularity is a key element of maintenance drug treatment; compliance is crucial for treatment success. Once- or twice-daily intake of a drug is always easier to comply with than regimens requiring more frequent dosing. Bronchodilating treatment was used as an example to illustrate how sustained duration of effect can be achieved by two different approaches: oral administration of the terbutaline prodrug bambuterol and inhalation of formoterol. Bioanalytical methods were employed to monitor the kinetic fate of bambuterol and formoterol in plasma, urine, or faeces. Generated terbutaline in plasma was used as a marker of effect for bambuterol. Established clinical laboratory tests were used to assess local and systemic effects of inhaled formoterol compared with salbutamol. Recommended doses of bambuterol, 10-20 mg once daily in adults, normally produced plasma concentrations of the active moiety terbutaline within therapeutically relevant limits. Dose proportionality for terbutaline makes dosing with bambuterol predictable. Compared with adults, children should be given higher doses than indicated by their lower body weight. Pharmacokinetic analysis indicated that absorption of bambuterol was slow and multi-phasic and that slow biotransformation to terbutaline occurred both presystemically and systemically. Systemically circulating formoterol was rapidly eliminated, the inactive (S;S)-formoterol more rapidly than the active (R;R)-formoterol. An inactive phenol glucuronide was the main metabolite, and a previously unknown sulphate metabolite was discovered. Duration of systemically mediated cardiovascular or metabolic side-effects of inhaled formoterol seemed not to differ from those of an inhaled systemically equieffective dose of salbutamol. There was a trend suggesting that the magnitude of systemic side-effects may be less pronounced after inhalation of formoterol compared with a locally equieffective dose of inhaled salbutamol.

Both approaches to sustaining stimulation of β2-adrenoceptors have their pros and cons. Bambuterol can be dosed orally once daily, but full effect is reached slowly. The effect of formoterol is reached within a few minutes, but administration must occur via the lungs, often twice daily. Both treatments, however, give 24-h symptom relief during regular treatment.

Johan Rosenborg, Department of Pharmaceutical Biosciences, Division of Pharmacokinetics and Drug Therapy, BMC, Box 591, SE-751 24 Uppsala, Sweden and Experimental Medicine, AstraZeneca R&D Lund, SE-221 87 Lund, Sweden

 Johan Rosenborg 2001

ISSN 0282-7484 ISBN 91-554-5099-7

Printed in Sweden by Lindbergs Grafiska HB, Uppsala 2001 PAPERS DISCUSSED

This thesis is based on the following papers, which will be referred to in the text by their Roman numerals.

I. Rosenborg J, Larsson P, Nyberg L. Pharmacokinetics of bambuterol during oral administration of plain tablets and solution to healthy adults. Br J Clin Pharmacol 2000;49:199-206. II. Ahlström H, Alvero J, Alvero R, et al. Pharmacokinetics of bambuterol during oral administration to asthmatic children. Br J Clin Pharmacol 1999;48:299-308. III. Rosenborg J, Larsson P, Tegnér K, Hallström G. Mass balance and metabolism of 3H-formoterol in healthy men after combined i.v. and oral administration - mimicking inhalation. Drug Metab Dispos 1999;27:1104-1116. IV. Rosenborg J, Bengtsson T, Larsson P, Blomgren A, Persson G, Lötvall J. Relative systemic dose potency and tolerability of inhaled formoterol and salbutamol in healthy subjects and asthmatics. Eur J Clin Pharmacol 2000;56:363-370. V. Rosenborg J, Larsson P, Rott Z, Böcskei C, Poczi M, Juhász G. Assessment of a relative therapeutic index between inhaled formoterol and salbutamol in asthma patients. Submitted 2001. ABBREVIATIONS USED IN FRAME STORY

Ae Amount excreted in urine during a collection interval ANOVA Analysis of variance AUC Area under the curve of plasma concentration vs time CLA Apparent clearance after oral administration (dependent also on bioavailability) CLR Renal clearance ESI Electrospray ionization FEV1 Forced expiratory volume (L) in the first second 3 H2O Tritiated water F Formoterol FG1 Phenol glucuronide of formoterol FG2 Benzyl glucuronide of formoterol FS Sulphate of formoterol S-K+ Serum concentration of the potassium ion LC Liquid chromatography LOQ Limit of quantification Met1 O-demethylated formoterol Met1G1 Phenol glucuronide 1 of O-demethylated formoterol Met1G2 Phenol glucuronide 2 of O-demethylated formoterol Met2S Sulphate of deformylated formoterol MS Mass spectrometry Q-Tc Q-T interval corrected for heart rate (R;R)/(S;S) Ratio of formoterol enantiomers in urine T½ Terminal half-life Total Total radioactivity: the sum of formoterol plus metabolites and tritiated water excreted in urine and faeces XTOT Quantified unidentified metabolites of formoterol CONTENTS INTRODUCTION...... 7

BASIC PHARMACOLOGY OF β2-AGONISTS...... 7

STRUCTURE, ACTIVITY, AND SELECTIVITY ...... 7 PULMONARY PHARMACOLOGY ...... 9 EXTRAPULMONARY PHARMACOLOGY ...... 9 METABOLIC EFFECTS...... 10 THE RELATION BETWEEN PHARMACOKINETICS AND PHARMACODYNAMICS ...... 10

PHARMACOKINETIC QUALITIES OF LONG-ACTING β2-AGONISTS...... 10 PROLONGED DURATION OF BRONCHODILATION – PREVIOUS RESULTS...... 12

PROLONGED SYSTEMICALLY-MEDIATED EFFECT OF THE SHORT-ACTING β2-AGONIST TERBUTALINE USING PRODRUGS ...... 12 PHARMACOKINETIC BASIS OF PROLONGED TOPICALLY-MEDIATED EFFECT ...... 15 PRESENT INVESTIGATION...... 17

OBJECTIVES...... 17 METHODS ...... 17 Pharmacokinetics of bambuterol...... 17 Fitting a pharmacokinetic model to previous data on bambuterol...... 17 Study designs ...... 18 Bioanalysis...... 19 Calculations and primary statistical analyses...... 20 Pharmacokinetics and effects of formoterol ...... 21 Study designs ...... 21 Effect measurements and bioanalysis...... 22 Calculations and primary statistical analyses...... 23 RESULTS...... 24 Pharmacokinetic modelling of previous data on bambuterol...... 24 Pharmacokinetics of bambuterol tablets and solution (Papers I and II)...... 25 Mass balance and metabolism of 3H-formoterol in healthy men (Paper III) ...... 28 Relative systemic dose potency and tolerability of inhaled formoterol and salbutamol in healthy subjects and asthmatics (Paper IV) ...... 29 Assessment of a relative therapeutic index between inhaled formoterol and salbutamol in asthma patients (Paper V)...... 29 DISCUSSION...... 31 Clinical pharmacokinetics of bambuterol...... 31 Clinical pharmacokinetics of formoterol...... 32 Concluding remarks...... 33 ACKNOWLEDGEMENTS...... 35

REFERENCES...... 36

APPENDIX ...... 40

INTRODUCTION

Pharmacologically mediated bronchodilatation is an important element in the treatment of asthma. The effect is primarily obtained with β2-adrenoceptor stimulants (“β2-agonists” in the following), normally administered by inhalation, but also by oral, subcutaneous, or intravenous routes. Inhaled short-acting β2-agonists are used primarily as needed. Long-acting β2-agonists are an alternative to increasing the maintenance dose of inhaled steroids (1). Three long-acting β2-agonists are available on the market: the terbutaline prodrug bambuterol for oral administration, formoterol and salmeterol for inhalation. The dosage forms of all three compounds comprise racemic mixtures of one active and one inactive enantiomer.

Bambuterol is an amphiphilic bisdimethylcarbamate of the β2-agonist terbutaline. Duration of action is prolonged because a therapeutically relevant plasma concentration of terbutaline, the active metabolite, is sustained. Formoterol and salmeterol are lipophilic drugs with high affinity for airway tissues. Maintained high lung concentrations of these drugs are probably important for their prolonged bronchodilation after inhalation (2). The clinical-pharmacokinetic aspects of prolonged effect duration are addressed in this thesis. Insights into the benefits conferred by different routes of administration are discussed on the basis of clinical studies of bambuterol and formoterol.

Enantiomeric mixtures were normally employed to monitor the kinetic fate of bambuterol and formoterol in plasma, urine, or faeces. Pharmacodynamic variables were used as markers of local and systemic effects of formoterol. Plasma concentration of the enantiomeric mixture of terbutaline was used as a marker of the effect potential for bambuterol. This was considered acceptable since the ratio between plasma concentrations of active and inactive enantiomers of terbutaline generated from bambuterol varies little over time, particularly during repeated oral treatment with bambuterol (AstraZeneca, data on file).

BASIC PHARMACOLOGY OF β2-AGONISTS

Structure, activity, and selectivity

Noradrenaline and adrenaline are the two important endogenous compounds mediating β-adrenoceptor stimulation. β1-adrenoceptors are stimulated primarily by noradrenaline released from nerve endings, β2-adrenoceptors in blood vessels, the airways, and skeletal muscles by circulating adrenaline. Adrenaline, the unselective prototype (stimulates both α- and β-adrenoceptors) of all β-adrenoceptor stimulating compounds, is rapidly metabolized by catechol-O-methyl transferase (COMT) and monoamine oxidase (MAO) and therefore has a short duration of action.

Synthesized analogues of adrenaline have proven useful in the pharmacological treatment of asthma (Figure 1).

7 OH

HO CH CH2 NH CH3 OH * Adrenaline OH CH OH CH3 3

HOCH CH NH C CH

HO CH CH2 NH CH 2 3

OH * OH * CH CH3 Isoprenaline 3 Colterol

HO OH CH3 OH CH3

HOCH CH2 NH C CH3

CH CH2 NH C CH3 * OH * HOCH CH3 Terbutaline CH3 2 Salbutamol

H3C O

H3C O O

O N CH C OH CH3 C OH CH3

H3C H3C CH CH2 NH C CH3 CH CH2 NH C CH3

H3C O

H3C O

O * O * CH CH C Ibuterol CH3 N C Bambuterol 3 H C H3C 3

OH CH3

HO CH CH NH CH CH OCH **2 2 3 HC NH Formoterol

O

HO OH CH3

CH CH2 NH CH CH2 OH OH * * Fenoterol

OH

HO CH CH2 NH (CH2)6 O (CH ) * 2 4

HOCH 2 Salmeterol

Figure 1. Structural formulæ of β2-adrenoreceptor stimulators mentioned in the text. Adrenaline, the endogenous prototype of all β2-agonists, has low selectivity and short duration of action; Amine substituted adrenaline with slightly prolonged effect duration and increased selectivity: isoprenaline, the isopropyl derivative, giving increased β-adrenergic selectivity, and colterol, the tertiary butyl derivative, giving increased β2-adrenergic selectivity; Substituted colterol giving further prolonged effect duration and selectivity: terbutaline, the resorcinol derivative, and salbutamol, the 3-methylhydroxy derivative; β2- adrenergically inert prodrugs of terbutaline: ibuterol, the diisobutyryl ester, is rapidly bioactivated and bambuterol, the bisdimethylcarbamate, is slowly bioactivated; Formoterol, a structure analogue of the short-acting β2-agonist fenoterol, is long-acting per se after inhalation; Salmeterol, a structure analogue of the short-acting β2-agonist salbutamol, is also long-acting per se after inhalation. Asterisks indicate chiral centres.

The common objective in the development of such drugs was to achieve high β2- adrenoceptor selectivity, metabolic stability, and long duration of action. Retrospectively, two strategic trends are discernable, i.e., amine and phenyl substitution of adrenaline. The former reduced inactivation by MAO and the latter inactivation by COMT. Importantly, for the intended use in treatment of asthma, both these manipulations of adrenaline increased β-adrenoceptor selectivity. The isopropyl derivative of adrenaline, isoprenaline (isoproterenol), was shown to be β-adrenoceptor- selective, and the tertiary butyl derivative, colterol, was shown to be β2-adrenoceptor- selective (3).

Substituting the catechol structure of colterol with resorcinol gives the still more β2- adrenoceptor-selective agonist terbutaline (4), which is not a substrate for MAO or COMT. Substituting the catechol structure of colterol with saligenin results in salbutamol, which also has increased β2-adrenoceptor selectivity (5), and which, like terbutaline, is not a substrate for either MAO or COMT. Because of their relatively high metabolic stability, terbutaline and salbutamol can be systemically administered and

8 their duration of action is prolonged compared with the solely amine-substituted analogues of adrenaline. Prodrugs aimed at sustaining the effect of terbutaline and formoterol, which not needs chemical modification to achieve long duration, are discussed below.

Pulmonary pharmacology

β2-agonists are functional antagonists. Stimulation of β2-adrenoceptors, located in the cell membrane of smooth muscles in the airways, decreases tone, irrespective of the cause of the constriction, i.e., the calibre of constricted airways is increased. Forced expiratory volume during in the first second (FEV1) is used as a marker of bronchodilating effect throughout this presentation. The persisting bronchodilatating effect of β2-agonists under regular treatment was observed at an early stage {e.g. (6)}. A low inclination for development of tolerance may be explained by a large number of “spare receptors” or a high level of gene expression for the β2-adrenoceptor (7).

Parasympathetic stimulation of the airways via cholinergic nerves increases the tone of smooth muscles and consequently reduces airway calibre. Furthermore, constrictive reactions in the airways can be triggered, e.g. by allergens, cold air, and exercise. The inhaled dose of a bronchoconstrictor giving a certain response, e.g. the dose of methacholine giving a decrease in FEV1 of 20 %, is normally used as a measure of such reactivity.

Increased mucociliary clearance and inhibition of vascular leakage and mediator release from inflammatory cells are positive effects of β2-agonists but their clinical role is unclear (7).ADDIN

Extrapulmonary pharmacology

Skeletal muscle tremor, tachycardia and palpitations, increased systolic and decreased diastolic blood pressure are well known clinical side-effects of β2-agonists.

The tremorogenic action of β2-agonists is probably mediated via local β2-adrenoceptors on the plasma membrane of skeletal muscle cells (8).

The cardiovascular effects of adrenoceptor agonists are mediated via α- and β- adrenoceptors. Adrenaline has a positive inotropic and chronotropic effect, mediated primarily via β1-adrenoceptors. Cardiac β2-adrenoceptors are less frequent, but selective β2-agonists still have direct positive chrono- and inotropic effects. The clinical consequences are tachycardia/palpitations and increased systolic blood pressure. Diastolic blood pressure is decreased as a consequence of adrenergically mediated peripheral vasodilatation.

9 Metabolic effects

Reversible cellular uptake of potassium (9) and increased glycogenolysis in the liver (10) are typically seen metabolic effects of β2-agonists. β2-agonists are calorigenic and so increase oxygen consumption (11). Because the supply of oxygen to the muscles and capacity for glycogenesis are limited (10), plasma lactate may increase secondarily to the increase in plasma glucose. These side-effects of β2-agonists, e.g. serum potassium suppression {severe hypokalaemia may prolong the Q-T interval (12) and be arrhythmogenic (13)}, are usually reduced during repeated treatment, because extrapulmonary β2-adrenoceptors are down-regulated (14, 15).

The relation between pharmacokinetics and pharmacodynamics

Several pharmacodynamic predictions based on pharmacokinetics imply a direct cause- effect relationship between drug concentration and effect. This means that the effect lags negligibly behind the measured change in drug concentration. If there is a lag between concentration and effect, a pharmacokinetic-pharmacodynamic (PK/PD) model must be used that adequately takes the lag into account.

The systemically mediated bronchodilatating response to terbutaline in the average asthmatic adult has been found to be linearly correlated to systemic concentrations of 7-27 nmol/L {1.6-6 ng/mL (16-18)} and a similar relation was observed in children (19). Thus, well-treated asthmatic patients, adults and children relying on systemically administered terbutaline should usually have drug concentrations between 5 and 30 nmol/L. A concentration below this range is likely to be therapeutically inadequate, and appreciable side-effects are frequent at concentrations above the range.

There is also a cause-effect relationship between plasma concentration and effects of locally administered drugs. However, as topical effects may be elicited before any appreciable amount of drug has been absorbed into the systemic circulation, local concentrations would in this case be more relevant for prediction of therapeutic effect. Pharmacokinetic aspects of pulmonary effect of systemically administered and inhaled drugs are illustrated in Figure 2.

Pharmacokinetic qualities of long-acting β2-agonists

Any approach to achieving long duration of effect aims at a sustained effective concentration at the site of action. The perfect drug would be 100 % systemically bioavailable via the oral route and have high efficacy with respect to the desired effect only. Because duration of action is normally dose-dependent (20), sustained effect of a drug with such selectivity could be obtained simply by having a high enough dose (21).

10 β2-agonists are highly selective for β2-adrenoceptors but their selectivity cannot be exploited for lung targeting via the systemic circulation, since typical side-effects such as tremor and palpitations are mediated via receptors outside the lungs. Thus, insufficient duration of action of systemically-administered β2-agonists cannot be compensated for by increasing the dose, because high systemic peak concentrations make side-effects unacceptable. Any pharmacokinetic method employed to sustain the bronchodilating effect should prevent the systemic drug concentration from reaching high levels. For oral administration, this can potentially be achieved by using sustained- release formulations or prodrugs. Oral administration Inhalation

Csyst→ Desired effect Clu ng → Desired effect ⇒ speed of onset ↑ ⇒ potency ↑ ⇒ selectivity ↑

Effector Effector organ organ

Side- Side- effects effects

Figure 2. Systemic concentration of a β2-agonist is an important determinant for bronchodilatating effect after oral administration whereas local concentration in the lungs would be more relevant after inhalation

The rationale for using inhaled β2-agonists is improvement of bronchodilating dose potency and a more rapid onset of effect compared with administration by the oral route. A consequence of improved topical potency after inhalation is an improved selectivity for the desired effect, which means that a relatively higher dose can be administered to the site of action by inhalation than by oral administration. Thus, the duration of action after systemic administration can in principle be prolonged by topical administration to the lungs. The prerequisite of this approach, however, is that the β2- agonist has a high affinity for lung tissues (or is slowly released and absorbed from the airway lumen), so that the drug will be retained in the lung after inhalation. Otherwise, the effect will rapidly disappear as the concentration difference between airways and the systemic circulation decreases with time towards the ratio obtained after oral therapy.

One of the two inhaled long-acting β2-agonists, salmeterol, is the end product of deliberate development of a β2-agonist with longer duration of action after inhalation than salbutamol. Lipophilic analogues of salbutamol were designed, which would more readily interact hydrophobically with the β2-adrenoceptor and thereby prolong duration of action. The concept worked although it became clear that increased receptor affinity was not the only explanation for the sustained effect of salmeterol (22). The sustained action of inhaled formoterol, originally developed for oral administration (23), was discovered by chance (24).

11 PROLONGED DURATION OF BRONCHODILATION – PREVIOUS RESULTS

Prolonged systemically-mediated effect of the short-acting β2-agonist terbutaline using prodrugs

A regular three-times-daily dosing regimen of conventional terbutaline sulphate (“terbutaline” in the following) tablets 5 mg is needed to give 24-hour symptom relief, indicating that concentrations at sites of action soon become subtherapeutic despite a lengthy terminal half-life of 17 h (25). Terbutaline’s duration of bronchodilation cannot be prolonged by a change to inhalation, probably because retention in the airways is poor, implying that the concentration in the lungs rapidly reaches a subtherapeutic level.

The original rationale for developing a systemically administered prodrug of terbutaline was to increase oral bioavailability. Initial attempts were made with acyl-substituted diesters of the parent drug. The long-chain carboxylic acids made the prodrugs more lipohilic, and the idea was that this would facilitate gastrointestinal absorption compared with terbutaline. Additionally, presystemic metabolic protection against sulphate conjugation was expected, leading to increased oral bioavailability of the parent drug (26). The diesters of terbutaline are bioactivated by pseudocholinesterase (EC 3.1.1.8) and unspecific esterases. In-vitro data suggested that the bulkier the acyl substituent the slower the bioactivation; it might therefore be possible to use a diester as an inner systemically circulating depot of terbutaline (27).

The concept above was tested with ibuterol, the diisobutyryl ester of terbutaline (Figure 1). Rate of absorption and extent of bioavailability were indeed improved with ibuterol, but a drawback was the shorter duration of action compared with terbutaline (28).

Bambuterol is the bisdimethylcarbamate prodrug of terbutaline. The adrenergically inactive bambuterol (29) has in itself a high affinity for lung tissue, but to become effective it must be administered systemically, preferably by the oral route. The reason is that biotransformation of bambuterol to terbutaline predominantly takes place outside the lungs via hydrolysis and oxidation, mediated by pseudocholinesterase and, most likely, by cytochrome P450-dependent oxidases, respectively (30-32). Intact bambuterol, the monocarbamate of bambuterol, and hydroxylated derivatives of bambuterol and the monocarbamate have been found in man after bambuterol administration (AstraZeneca, data on file). The monocarbamate is generated via both the oxidative and hydrolytic pathways, whereas the hydroxylated metabolites are generated only during oxidative degradation of bambuterol (Figure 3). Demethylated bambuterol and demethylated monocarbamate, formed from the hydroxylated compounds, are unstable and are rapidly hydrolyzed to the monocarbamate or terbutaline. Bambuterol, in contrast to ibuterol, inhibits its own hydrolysis, making it a potential inner depot from which terbutaline is slowly generated (33).

12 H3C O

N C O OH CH3

H3C CH CH2 NH C CH3

H3C O O * N C Bambuterol CH3 H3C

HOCH2 O

O N C OH OH CH3 OH CH3

H3C

CH CH2 NH C CH3 CH CH2 NH C CH3

H3C O H3C O

O * O * N C CH3 N C CH3

H3C H3C

HOCH 2 O

O N C OH OH CH3 OH CH3

H3C CH

CH CH2 NH C CH3 O CH2 NH C CH3

HOCH 2 O HOCH 2

O O * * N C CH3 N C CH3

H3C H3C

HO OH CH3

CH CH2 NH C CH3 OH * Terbutaline CH3

Figure 3. Biotransformation of the prodrug bambuterol to terbutaline via hydrolytic and oxidative pathways, mediated by pseudocholinesterase and probably by cytochrome P450-dependent oxidases, respectively. The dotted arrows indicate hydrolysis catalysed by pseudocholinesterase and the solid arrows indicate other mainly oxidation-driven pathways. Adapted from Tunek et al. (30), Svensson & Tunek (31), and Lindberg et al. (32). Short-lived demethylated metabolites produced from the hydroxylated compounds are not shown in the metabolic scheme. Asterisks indicate chiral centres.

The basal pharmacokinetics of bambuterol and generated terbutaline have been evaluated in healthy subjects, by use of data after intravenous administration of bambuterol hydrochloride (“bambuterol” in the following) 30 µg per kg body weight (on a molar basis equivalent to about 20 µg of terbutaline) and terbutaline 7 µg/kg, and after single and once-daily repeated oral dosing of bambuterol 0.270 mg/kg (34). The intravenous doses were given in two equal parts separated by one hour. Mean clearance of bambuterol, 1.25 L/min, was five times higher than the clearance of terbutaline, 0.23 L/min, but mean volume of distribution, 1.6 L/kg, was similar. The AUC for terbutaline was virtually the same after intravenous bambuterol and terbutaline, giving a mean bioavailability of terbutaline of 36 % after intravenous bambuterol administration.

The mean terminal half-life of bambuterol was 2.6 h after intravenous but 12 h after oral administration, implying rate-limiting gastrointestinal absorption.

The plasma concentration of intravenously infused terbutaline rapidly decreased from a very high level post-dosing to a low level during the terminal phase of elimination. The concentration of terbutaline slowly reached a lower peak after intravenous bambuterol, then decreased monophasically at a higher level than after the intravenous

13 administration of terbutaline as such. This obvious benefit was fortified after oral administration of bambuterol (Figure 4).

100.0

10.0

1.0

Plasma concentration (nmol/L) 0.1 010203040 Time after administration (h)

Figure 4. Single-dose plasma concentration vs. time of terbutaline (open squares after intravenous terbutaline; open circles after intravenous bambuterol; open diamonds after oral bambuterol) {adapted from Nyberg et al (34)}. Dotted horizontal lines indicate the alleged target range for effective concentrations of terbutaline (cf. text).

Mean bioavailability of generated terbutaline was found to be 10 % after oral bambuterol. Mean absorption of bambuterol after oral administration is about 20 % (AstraZeneca data on file). Thus, approximately half the absorbed dose is ultimately transformed into active agonist, implying that first-pass metabolism makes a considerable contribution to the bioavailability of terbutaline after oral administration of bambuterol.

The mean plasma concentration peak of terbutaline during the once-daily oral regimen was 15 nmol/L and the minimum 8 nmol/L, i.e., within the therapeutic target range for systemically-mediated bronchodilation, as discussed above. Similar minimum concentrations were found in adult asthmatic patients given the same dosing regimen (35, 36).

The pharmacokinetic properties of bambuterol make it suitable for oral once-daily dosage (37), maintaining bronchodilatation around the clock (36, 38-40). The clinical efficacy of bambuterol 10 mg once daily was approximately equieffective with that of a controlled-release preparation of terbutaline 5 mg twice daily (41), whereas bambuterol 20 mg once daily was approximately equieffective with plain terbutaline tablets 5 mg three times daily (42). The side-effects with once-daily regimens of bambuterol 10-20 mg appear to be less pronounced than with equi- or less effective oral reference treatments with terbutaline and salbutamol (41-44). Regular oral once-daily regimens of bambuterol compete favourably with regular twice-daily inhalation of salmeterol (45, 46).

14 The recommended treatment of asthmatic adults is regular once-daily administration of bambuterol tablets or solution 10 or 20 mg. The bronchodilating effect of bambuterol was little improved by increasing the dose to more than 20 mg/day (≈0.27 mg/kg/day), but side-effects increased (35). Dosing in the evening is preferable in patients with nocturnal asthma, in whom mean FEV1 in the early morning was improved by as much as 29 % over placebo during repeated treatment with bambuterol tablets 20 mg (39).

Previous studies of bambuterol had given valuable information for proper dosing in adults, but there was still lack of knowledge. This dissertation has focused on gaps regarding possible mechanisms of bioactivation, dose linearity, and dosage in children.

Pharmacokinetic basis of prolonged topically-mediated effect

The physico-chemical characteristics may be important determinants of onset and duration of effect for β2-agonists. Some possible microkinetic implications of the physico-chemical differences between short- and long-acting β2-agonists are given in Figure 5 (2). Hydrophilic β2-agonists such as terbutaline, salbutamol, and fenoterol, are distributed in the extracellular aqueous biophase of the whole body – within as well as outside the airways. Probably, the extracellular concentration of these drugs therefore rapidly attains high levels at the site of action in the airways after inhalation but then rapidly decline. Consequently, onset of action is rapid but duration of effect short.

Extracellular space

Lipid bilayer

Salbutamol Formoterol Salmeterol Hydrophilic Intermediate Lipophilic Short duration Long duration Long duration Fast onset Fast onset Slow onset

Figure 5. Physico-chemical characteristics may explain why onset and duration of effect vary

between different β2-agonists. The hydrophilic salbutamol, which rapidly diffuses to the receptor and is distributed in the body via the aqueous biophase, has a rapid onset but short duration of effect. The lipophilic salmeterol, which is primarily distributed from the aqueous to a lipid biophase, such as the lipid bilayer of the cell membrane, has a slow onset but long duration of action. Formoterol, which has intermediary lipophilicity, would after inhalation be expected to have pharmacodynamic properties somewhere in between those of salbutamol and salmeterol. Small arrows indicate distribution equilibrium of drug between aqueous and lipid biophases, and large black arrows indicate major movement and interactions. The supposed

fates of β2-agonists near the receptor are adapted from Anderson et al. (2).

15 A lipophilic β2-agonist such as salmeterol, however, is primarily distributed from the aqueous to a more lipid biophase, such as the lipid bilayer of the cell membrane, where it can be retained. Probably, a pool of salmeterol is slowly redistributed from the lipid biophase to the receptor either via the extracellular fluid or via the lipid bilayer. Thus, inhaled salmeterol would attain a clinically relevant concentration at the site of action in the airways slowly, but the effective concentration would then be maintained for a longer period of time than for an inhaled hydrophilic (short-acting) analogue. This could explain both the slow onset and long duration of action of inhaled salmeterol. It has also been suggested that the sustained action of salmeterol could depend on specific hydrophobic interactions with molecular regions near the actual β2-adrenoceptor (47). Formoterol, which has intermediary lipophilicity, would after inhalation be expected to have pharmacodynamic properties somewhere in between those of salbutamol and salmeterol. Löfdahl et al. (24) first showed, in line with this hypothesis, that inhaled formoterol in fact has both a rapid onset and long duration of effect (Figure 6).

Bronchodilation (%) 100

80 Inhaled 60 Oral 40

20 0

-20 0 60 120 180 240 300 360 420 480 Time (min)

Figure 6. Bronchodilation of formoterol (% of maximum

increase in FEV1) is greater and more sustained after inhalation (6 µg) than after oral administration (80 µg). At 8 h post-dosing, 1.25 mg of terbutaline sulphate was inhaled. Adapted from Löfdahl et al. (24).

The bronchodilating effect of formoterol (formoterol as dose = formoterol fumarate dihydrate) gradually increased within the dose range 4.5–18 µg after inhalation via Turbuhaler (48). Doses of formoterol up to 18 µg, delivered via Turbuhaler, are well tolerated, and 36 µg is considered clinically safe although serum potassium is occasionally low and the incidence of tremor is higher than with lower doses (Data on file, AstraZeneca R&D). The recommended dose of inhaled formoterol for regular treatment of asthmatic adults is therefore 4.5–18 µg once or twice daily via Turbuhaler. It can be mentioned as comparison, that formoterol Turbuhaler 9 µg is approximately equieffective with salmeterol xinafoate Diskhaler 50 µg both with respect to its desired bronchodilating effect and to its systemic side-effects (48, 49).

16 Basic pharmacokinetics of inhaled formoterol had been investigated before, but little was known about degradation of formoterol and the fate of degradation products, particularly in man after inhalation. Thus, in this dissertation, mass balance and metabolism have been studied after a dosing regimen specifically designed to mimic inhalation. Furthermore, although previous studies had shown the long-acting bronchodilating potential of inhaled formoterol, less was known about the relation between desirable and undesirable effects. Therefore, it was judged to be a safety interest to further study, in comparison with a standard short-acting alternative, the systemic effects of formoterol and the relation of side-effects to bronchodilation.

PRESENT INVESTIGATION

Objectives

1. To test a tentative pharmacokinetic model for bambuterol by use of previous data. 2. To investigate the basic pharmacokinetics of long-acting β2-agonists, and particularly regarding: • The proportionality between plasma exposure and dose during treatment with bambuterol in adults and children • The metabolism of formoterol in man 3. To investigate if an effective β2-agonist concentration can be maintained by means of systemic once-daily administration of bambuterol to different populations. 4. To assess the systemic effects of inhaled, therapeutically relevant formoterol doses, particularly regarding the duration of side-effects and the relation of side-effects to bronchodilation in comparison with a short-acting analogue.

Methods

Pharmacokinetics of bambuterol

Fitting a pharmacokinetic model to previous data on bambuterol

A pharmacokinetic model was fitted to plasma concentrations of enantiomeric mixtures of bambuterol and generated terbutaline to further explore previous data presented by Nyberg et al. (34). The model took into account the following features revealed in the non-parametric analysis: 1) the absorption of bambuterol is multi-phasic; 2) terbutaline is slowly generated from systemically circulating bambuterol via a pool of intermediate metabolites; and 3) intestinal absorption of bambuterol is slow and further levels out the plasma concentration profile of generated terbutaline. Figure 7 shows the integrated model that was fitted to plasma concentration data after intravenous and oral

17 Compartment 2: Compartment 5: peripheral peripheral compartment for compartment for bambuterol terbutaline

K45 K5 4 K1 2 K2 1

Compartment 8: Compartment 10: Presystemic depot intestinal depot K8 1 Compartment 7: K1 04 compartment for systemic pool of compartment for bambuterol Compartment 1: intermediate Compartment 4: bambuterol transformed F8 K7 4 F10 plasma bambuterol K1 7 metabolites plasma terbutaline to terbutaline in first VC VC pass (K104=K91) K9 1 K1 0 K4 0 Compartment 9: F9 intestinal depot K4 6 compartment for K3 1 K1 3 K6 4 bambuterol

Compartment 3: Compartment 6: peripheral peripheral compartment for compartment for bambuterol terbutaline

Figure 7. Integrated pharmacokinetic model for enantiomeric mixtures of bambuterol and

terbutaline. Three-compartment models with the same central volume of distribution (Vc) were used to describe systemic fates and a biphasic model to describe gastrointestinal absorption of bambuterol. A pool, tentatively consisting of intermediary metabolites, was used in the model to describe systemic biotransformation of bambuterol to terbutaline. Presystemic generation of terbutaline was set to be rate-limited by the slow gastrointestinal rate of absorption of bambuterol (K91=K104). administration of bambuterol and after an intravenous reference dose of terbutaline (cf. the section “Prolonged systemically mediated effect of the short-acting β2-agonist terbutaline using prodrugs”).

Study designs

Pharmacokinetic aspects of dosing the prodrug bambuterol were investigated in healthy adult subjects regarding time to steady state, dose proportionality, intraindividual variability, and bioequivalence between tablets and solution (Paper I), and in asthmatic children regarding dose proportionality, dosing frequency, and ethnic differences (Paper II). The recommended maintenance oral doses of bambuterol in Caucasian adults, 10 mg or 20 mg once daily, plus a supratherapeutic dose of 30 mg were investigated in the healthy subjects. The recommended doses of 10 mg or 20 mg constituted starting points also for the pharmacokinetic survey of bambuterol in children. Based on the initial results obtained in school children, regimens of reduced doses were investigated in pre- school children. Enantiomeric mixtures of bambuterol and terbutaline were used to monitor the kinetic fate in plasma and urine. The plasma concentration of enantiomeric mixture of terbutaline was used as a marker of the effect potential of bambuterol.

Paper I: Twenty-six healthy Caucasian subjects were included and 23 (12 women) completed an open, randomized, crossover study. Bambuterol tablets, 10, 20, and 10+20 mg, and a solution, 20 mg (dose given as bambuterol hydrochloride), were given

18 orally once daily for 2 weeks at about 7 p.m. Plasma concentrations and urinary recoveries of bambuterol and terbutaline were measured after single doses and during repeated treatments.

Paper II: Forty-eight asthmatic children in four different studies each completed two double-blind bambuterol 1- or 2-week treatments (daily doses of bambuterol hydrochloride): 12 pre-school (5 mg×2 vs 10 mg) and 12 school (10 mg vs 20 mg) Caucasians plus 12 pre-school (2.5 mg vs 5 mg) and 12 school Orientals (10 mg vs 20 mg). Plasma concentrations and urinary recoveries of bambuterol and terbutaline were measured during repeated treatments.

Demographic data of subjects and patients participating in the studies with bambuterol are summarized in Table 1.

Table 1: Demographic data on subjects and patients in prospective bambuterol studies

Paper Ethnic Group Number of Bambuterol Demographics Origin Subjects Hydrochloride Variable Mean Min Median Max (male/female) (mg/day) Healthy adult subjects I Caucasian Healthy 141/12 10, 20, 30 Age (years) 29 20 25 56 Adults Weight (kg) 67 46 69 81 Asthmatic children II Caucasian Pre-school2 8/4 5×23, 10 Age (years) 4.8 3 5 6 Asthmatics Weight (kg) 20.5 15 20 25 School4 9/3 10, 20 Age (years) 9.8 7 10 13 Asthmatics Weight (kg) 33.1 26 34 45 II Oriental Pre-school3 8/4 2.5, 5 Age (years) 3.5 2 3 5 Children Weight (kg) 15.3 11 14 25 School4 10/2 10, 20 Age (years) 7.7 6 7 10 Asthmatics Weight (kg) 23.6 16 21 44

1Two men who discontinued after one treatment were replaced. One man discontinued after two treatments and was not replaced 2Pre-school children were given bambuterol solution 3Administered twice daily 4School children were given bambuterol tablets

Bioanalysis

Bambuterol was analysed using non-chiral gas chromatography plus mass spectrometry (50). At the limits of quantification (LOQ), 1 nmol/L in plasma and 8 nmol/L in urine, the within-day coefficient of variation (CV) was 4% and 5%, respectively.

Terbutaline was analysed by use of non-chiral coupled-column high performance liquid chromatography with electrochemical detection, according to a modification of a 19 previously described method (51). At 8 nmol/L in plasma, the CV was 2.1% within a day and 3.6% between days. The applied LOQ was 4 nmol/L. The same method was used for terbutaline in urine. The assay calibration curve was then extended up to 5000 nmol/L to allow for measurement of the higher concentrations in urine. At 80 nmol/L (the LOQ) the within-day variation was 3.7% in the urine analysis.

Calculations and primary statistical analyses

A non-linear mixed effect model was used to evaluate previous data on bambuterol parametrically. The analysis was performed using the ADVAN5 subroutine and conventional first order estimation method in the NONMEM V 1.1 package (52). The final model was selected on the basis of the objective function value and graphical analysis of predicted vs observed values and residual plots. Parameters describing elimination of bambuterol and terbutaline (K10, K40), systemic irreversible biotransformation from bambuterol to terbutaline (K17, K74), oral bioavailability of bambuterol (F8, F9), and the fraction of orally administered bambuterol that is presystemically transformed to terbutaline (F10) were assumed to be normally distributed random variables within the group of subjects; other parameters were not assumed to vary between subjects. Intraindividual variability in pharmacokinetic parameters was not regarded. The model was fitted to log-transformed data.

The elimination rate constants of bambuterol and terbutaline were calculated from urinary excretion data, ln (dAe/dt) regressed on midpoint of collection period. Terminal half-life (T½) is ln2 divided by the elimination rate constant. Area under the curve of plasma concentration vs time (AUC) was calculated using the trapezoidal rule. Extrapolation of AUC to infinity was done assuming monoexponential decline after the last measurable concentration according to the estimated terminal half-life. Apparent clearance after oral administration, CLA, and renal clearance, CLR, were calculated as dose and urinary recovery, respectively, divided by AUC.

Treatments were statistically compared and described using multiplicative or additive (T½) ANOVA models normally with the factors subject, period, and treatment, plus dose number in the healthy adults.

Time to steady state for terbutaline was investigated in the healthy adults by comparing AUC0-24h after dose 15 and dose 8 of bambuterol tablets 20 mg and by comparing Ae0-24h after dose 15 and doses 8-14 of bambuterol tablets and solution 20 mg - it was assumed that bambuterol had reached steady state after dose 8 and terbutaline after dose 15 (Paper I).

Dose proportionality was addressed by use of AUC/dose or Ae/dose (Papers I and II).

Steady-state intraindividual variability in systemic exposure was assessed indirectly from Ae0-24h after doses 8-15 of bambuterol tablets and solution 20 mg, and directly from AUC0-24h after dose 8 and dose 15 of bambuterol tablets 20 mg - the assumption of this analysis was that steady state was reached at dose 8 (Paper I).

20 Bioequivalence of generated terbutaline between tablets and solution was evaluated by use of the ratios of AUC0-24h and Ae0-24h after dose no. 15 of bambuterol tablets and solution 20 mg. The criterion for bioequivalence was a 90% confidence interval contained within 80-125% (Paper I).

Pharmacokinetics and effects of formoterol

Study designs

Pharmacokinetic aspects of formoterol dosing and systemic dose response after inhalation were studied in healthy subjects (Papers III and IV). A modelling approach for assessing a relative therapeutic index – the relation between therapeutic ratios of drugs exerting similar effects – was developed comparing inhaled formoterol with inhaled salbutamol (Paper V). A supratherapeutic single or cumulative formoterol dose of 54 µg was used to investigate basic pharmacokinetics, metabolism, tolerability, and systemic potency. Single formoterol doses covering the recommended range of 4.5-18 µg via Turbuhaler, plus the supratherapeutic dose of 54 µg were used to relate systemic to local effects. The effects of formoterol were referred to those of inhaled salbutamol in doses up to 3600 µg via a pressurized metered dose inhaler. Enantiomeric mixtures were used to monitor the kinetic fate of formoterol in plasma, urine, and faeces (Papers III and IV). Terminal elimination was followed enantioselectively in urine (Paper IV). Pharmacodynamic variables were used as markers of local and systemic effects of formoterol (Papers IV and V).

Paper III: Mass balance and metabolism of 3H-formoterol were investigated in six healthy men in an open fashion. Simultaneous oral (mean dose 88.6 nmol, 49.3 MBq) and intravenous (mean dose 38.2 nmol, 21.4 MBq) doses of tritium-labelled formoterol were administered. The combination of these two administrations aimed at simulating the fate of inhaled formoterol. Total radioactivity was monitored for 24 h in blood plasma and for at least 4 days in urine and faeces.

Paper IV: Relative systemic dose potency and tolerability of inhaled formoterol and salbutamol were compared in β2-agonist naïve and accustomed subjects and patients, respectively, and pharmacokinetic enantioselectivity was described. Twelve healthy subjects completed three open study days, and eleven asthmatic patients completed four double-blind study days in randomized, placebo-controlled, and crossover fashion. The healthy subjects inhaled 13.5+13.5+27 µg of formoterol fumarate dihydrate (“formoterol”, Oxis) via Turbuhaler and 300+300+600 µg of salbutamol (Ventolin) via a pressurised metered dose inhaler (pMDI) on two separate days. The asthmatics, who were on formoterol 9 µg twice daily via Turbuhaler during the study, inhaled the same single doses of formoterol or salbutamol as the healthy subjects on two separate days, and on an additional day 900+900+1800 µg of salbutamol via pMDI. Doses were given cumulatively 30 min apart. Placebo was a day of no treatment in the healthy subjects and double dummies were used for the asthmatics. Cardiovascular and

21 metabolic effects were evaluated. Plasma concentrations and urinary excretion of formoterol were monitored in the healthy subjects.

Paper V: The bronchodilating and serum-potassium-suppressing effects of inhaled formoterol and salbutamol were measured in twenty-eight stable asthmatic patients who completed a double-blind, randomized crossover study. Baseline FEV1 (mean 2.08 L) was 49–93% of predicted and reversibility was 16–82% after inhalation of 200-400 µg salbutamol. Patients inhaled three single doses of formoterol (Oxis, delivered doses of 4.5, 18, and 54 µg) via Turbuhaler, two single doses of salbutamol (200 and 1800 µg) via a pressurised metered dose inhaler (pMDI), and placebo at intervals of 48 h or more. Maximum local and systemic effects were assessed. A classic sigmoid model of log- dose response was used to discriminate pharmacologically between formoterol and salbutamol. Relative local and systemic dose potencies and their ratio - the relative therapeutic index - were estimated simultaneously using a non-linear mixed effect modelling approach.

Demographic data of subjects and patients participating in the studies of formoterol are summarized in Table 2.

Table 2: Subjects and patients in formoterol studies

Paper Topic Number of Administration Formoterol Comparator Demographics

Subjects Route Fumarate•2 H2O Variable Mean Min Median Max (male/female) (µg) Healthy adult subjects III Pharmacokinetics, 6/0 I.V. + Oral 16.2 + 37.8 - Age (years) 50 40 54 63 Mass balance Weight (kg) 85 73 89 99 IV Tolerability, 6/6 Inhalation 13.5+13.5+27 Salbutamol Age (years) 23 18 23 28 Pharmacokinetics Weight (kg) 69 58 67 82 Asthmatic patients IV1 Tolerability 5/6 Inhalation 13.5+13.5+27 Salbutamol Age (years) 34 29 29 56 Weight (kg) 76 65 74 93 V Therapeutic 14/14 Inhalation 4.5, 18, 54 Salbutamol Age (years) 43 20 44 64 Ratio Weight (kg) 72 49 70 91

1Single dosing on top of regular inhalation of formoterol 9 µg twice daily via Turbuhaler

Effect measurements and bioanalysis

Pulse rate and blood pressure were measured after a 5-minute rest with subjects or patients in the sitting position using standard techniques. The electrocardiograph automatically generated Q-Tc, i.e., Q-T interval corrected for heart rate. Reported adverse events and vital sign measurements were used to evaluate clinical tolerability (Papers IV and V).

22 Clinical laboratory tests were performed in plasma or serum routinely prepared from venous blood samples. An ion-selective electrode was used to measure serum potassium (Papers IV and V). Reagent kits were used to measure plasma glucose and plasma lactate (Paper IV). The maximum coefficient of variation in analyses of quality control samples was 1.0% for serum potassium, 5.1% for plasma glucose, and 6.9% for plasma lactate. Pulse (bpm, beats per minute) was measured via the radial artery. Forced expiratory volume in the first second (FEV1) was measured with the patient sitting in an upright position, using a Vitalograph Alpha spirometer (Vitalograph Ltd., U.K.) according to well established standards defined by the American Thoracic Society (53).

3H-formoterol and metabolites were determined using LC techniques plus radiodetection, directly after centrifugation in urine and after sample work-up in blood plasma and faeces. Metabolites were identified in urine, sampled from two subjects, using LC-ESIMS/MS techniques.

Inhaled formoterol was determined in urine and plasma by use of non-chiral coupled- column liquid chromatography with electrochemical detection (Paper IV). The limit of quantification was 0.50 nmol/L in urine and 0.05 nmol/L in plasma. The between-day variation in authentic samples was 4.8% at 5.0 nmol/L of formoterol in urine and 5.4 % at 0.18 nmol/L of formoterol in plasma. The enantiomeric ratio of formoterol was determined after chiral separation on an α1-acid glycoprotein column with electrochemical detection. The range for determination of enantiomeric ratio was 0.50 - 2.0 and the lower limit for determination of the ratio was 1.0 nmol/L of the enantiomeric mixture of formoterol. The between-day variation in authentic samples was 3.5% at an enantiomeric ratio of 0.60 at 37 nmol/L of total formoterol.

Calculations and primary statistical analyses

The elimination rate constant of formoterol was calculated from urinary excretion data, ln (dAe/dt) regressed on midpoint of collection period. Terminal half-life (T½) is ln2 divided by the elimination rate constant. Area under the curve of plasma concentration vs time (AUC) was calculated using the trapezoidal rule.

Treatments were statistically compared with respect to maximum effects (Paper V) or measurements after cumulative dose increments (Paper IV) using ANOVA models with factors subject or patient, treatment, and period (Papers IV and V), plus dose number and the interaction between treatment and dose number at cumulative dosing (Paper IV). The pre-drug administration measurement (a measurement made in the morning before administration of study treatment) was used as covariate.

Formoterol and salbutamol were assumed to have similar modes of action, implying that their log-dose response curves would be parallel. The possible device-related impact on the relation between local and systemic effects was not regarded. Relative dose potencies were estimated on this principle as the horizontal shift between parallel lines fitted to adjusted means, plotted against logarithms of doses (Paper IV), or using a 23 bivariate non-linear mixed effect model fitted simultaneously to all individual measurements (Paper V). Fieller’s method (54) was used to calculate confidence intervals for the estimates of relative dose potency in Paper IV. The asymptotic normality of the estimates was used to calculate confidence intervals for relative dose potencies and relative therapeutic index in Paper V.

Results

Pharmacokinetic modelling of previous data on bambuterol The applied model adequately described the fate of bambuterol after intravenous and oral administration and terbutaline after intravenous administration as such. Predicted plasma levels of terbutaline generated from bambuterol fairly well agreed with observations, although the terminal rate of decline was slightly overestimated (Figure 8). The program code used in the model analysis and goodness of fit plots are given in the Appendix. Intravenous terbutaline Intravenous bambuterol 1000.0 1000.0

100.0 100.0

10.0 10.0

1.0 1.0

0.1 0.1 0 10 20 30 40 0 10 20 30 40 Oral bambuterol, single dose Oral bambuterol, 7 days once daily 100.0 100.0

10.0 10.0 Plasma concentration (nmol/L) concentration Plasma 1.0 1.0

0.1 0.1 0 10 20 30 40 144 154 164 174 184

Time after dosing initiation (h)

Figure 8. Plasma concentration vs. time for terbutaline (dotted dashed lines) and bambuterol (dotted lines) normalized to intravenous bambuterol hydrochloride 2042 µg, intravenous terbutaline sulphate 477 µg, and oral bambuterol hydrochloride 16.2 mg. Predictions (continuous lines for both bambuterol and terbutaline) are based on the model shown in Figure 7. Adapted from Nyberg et al (34).

The outcome of the model analysis is summarized in Table 3. Clearance of bambuterol, calculated to 0.96 L/min (K10 × VC), was considerably higher than the clearance of terbutaline, calculated to 0.23 L/min (K40 × VC). The total rate of elimination of bambuterol was modelled as K10 (elimination not generating terbutaline) plus K17 (irreversible biotransformation to terbutaline via a systemic intermediary pool).

24 The fraction of terbutaline systemically transformed from intravenously administered bambuterol was estimated at about 35 % using the values for these parameters {K17/(K17+K10)}. Furthermore, oral bioavailability was calculated to about 10 % for bambuterol (F8 + F9) and 9.5 % for terbutaline {F10 + (F8 + F9)× K17/(K17+K10)}. None of these values contradicted the outcome of the previous non-parametric analysis (34). The slowly absorbed fraction of bambuterol (F9) was estimated to be about twice the more rapidly absorbed one (F8).

Table 3: Pharmacokinetic model analysis of literature data (34).

Variable Parameter Mean SEM Bambuterol K10, h-1 5.07 0.396 K12, h-1 5.05 0.744 K21, h-1 1.64 0.112 K13, h-1 2.92 0.560 K31, h-1 0.352 0.0344 K17, h-1 2.78 0.209 K81, h-1 0.474 0.110 F8 0.0386 0.00898 F9 0.0658 0.00900 Terbutaline K40, h-1 1.22 0.100 K45, h-1 0.882 0.114 K54, h-1 0.0924 0.0111 K46, h-1 2.39 0.327 K64, h-1 1.51 0.228 K74, h-1 0.311 0.0451 F10 0.0578 0.00408

Bambuterol/Terbutaline VC, L 11.4 0.646 K91;K104, h-1 0.0477 0.00583 Interindividual variability was considered for underlined parameters

Pharmacokinetics of bambuterol tablets and solution (Papers I and II)

Mean plasma concentrations of bambuterol and its active metabolite terbutaline after administration of bambuterol tablets, 10 and 30 mg to healthy adults, are shown in Figure 9. The initial absorption rate could not be assessed directly, but it was faster than during a second phase where absorption was rate-limiting for elimination, resulting in a mean terminal half-life of 16 h. The mean terminal half-life of terbutaline was 22 h and steady-state was reached within one week of bambuterol treatment.

Steady-state plasma exposure (AUC0-24h) of bambuterol was not linearly correlated to dose, but the overall pharmacokinetics of terbutaline indicated dose linearity. The mean minimum plasma concentration of terbutaline was 5–17 nmol/L within the studied range of bambuterol doses, and the corresponding mean maximum 12–31 nmol/L. Steady- state for terbutaline was reached after 8 days. Intraindividual variation between day 8 and day 15 in terbutaline AUC0-24h, 15% with the tablets, was half that of bambuterol. Urine data indicated that intraindividual variability was slightly lower with the solution.

25 Tablets were bioequivalent with the solution regarding terbutaline (90% confidence interval: 87-100%).

As in adults, plasma exposure of terbutaline but not of bambuterol was linearly related to the dose of bambuterol in asthmatic children. Mean plasma concentrations of generated terbutaline during repeated dosing of bambuterol are shown in Figure 10. Bambuterol Pre-school children School children 50 50 50 50

40 40 40 40 Caucasian

30 30 30 30

20 20 20 20

10 10 10 10

0 0 0 0 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 Terbutaline 50 50 50 50

40 40 Oriental 40 40

30 30 30 30

20 20 20 20

10 10 10 10 Mean concentration plasma (nmol/L) Mean plasma concentration (nmol/L) concentration plasma Mean 0 0 0 0 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 Time after administration of bambuterol (h) Time after administration of bambuterol (h)

Figure 9. Mean plasma concentrations of Figure 10. Mean plasma concentration of bambuterol and terbutaline after oral terbutaline after oral administration of administration of bambuterol tablets 10 and 30 bambuterol to Caucasian and Oriental children. mg (left panels) and tablets/solution 20 mg Dotted-line: 2.5 mg once daily; Dotted-dashed (right panels). Dashed lines are concentrations lines: 5 mg once (Oriental pre-school) or twice after single doses and continuous lines are (Caucasian pre-school) daily; Dashed lines: 10 concentrations at steady state. mg once daily; Continuous lines: 20 mg once daily.

Twice-daily dosing (2×AUC0-12h) could not be shown to differ from once-daily dosing (AUC0-24h) in the pre-school Caucasians, mean AUC for terbutaline being 128 and 242 h×nmol/L (5 mg/12h; 10 mg/24h). With once-daily dosing, terbutaline AUC was 213 and 406 h×nmol/L in the Caucasian school children (10 mg; 20 mg), 87.4 and 202 h×nmol/L in the Oriental pre-school children (2.5 mg; 5 mg), and 356 and 640 h×nmol/L in the Oriental school children (10 mg; 20 mg). Oriental school children had higher plasma concentrations of bambuterol and terbutaline than Caucasian school children. The strictly ethnical implication of the difference could not be elucidated, because demographic data were not perfectly matched. Terbutaline AUC was only moderately increased in the Caucasian school children compared with Caucasian adults during administration of bambuterol tablets 10 and 20 mg. With a once-daily dose of 10 or 20 mg, the increase was more pronounced in some pre-school Caucasians and generally in Oriental children. The highest concentration of terbutaline, 58 nmol/L, was seen in an Oriental school child after a 20-mg dose. The range of mean plasma concentrations for the applied regimens was otherwise fairly well within the previously discussed range of 5–30 nmol/L (cf. section “The relation between pharmacokinetics and dynamics”), except in the pre-school Orientals after 2.5 mg, where very low concentrations were found.

26 Clearance of both bambuterol and terbutaline was related to body weight. Individual values in adults and children are plotted in Figure 11. Note that the relation is blurred for bambuterol because of the non-linear pharmacokinetics. Body weight correction levels out much of the clearance difference for bambuterol between children and adults, whereas such an adjustment was not an effective means to scale clearance of terbutaline (Figure 12); clearance of terbutaline corrected for body weight was higher in the children, particularly in those weighing less than 40 kg (Table 4). Dosing according to body weight could therefore lead to undertreatment of the children.

Bambuterol Terbutaline 6000 360 Figure 11. Individual absolute

90000 900 320 5000 apparent (filled circles) and ) 80000 800 ) 280 70000 700 4000 240 renal (open circles) clearance of 60000 600 200 50000 500 3000 bambuterol and generated 160 40000 400 2000 120 terbutaline during oral 30000 300 Renal clearance (mL/min) Renal clearance (mL/min)

Apparent clearance (mL/min 80 20000 200 Apparent clearance(mL/min 1000 administration of bambuterol in 10000 100 40

0 0 0 0 children and adults 0 20406080100 0 20406080100 Weight (kg) Weight (kg)

Bambuterol Terbutaline 1800 18 250 15 Figure 12. Individual body 1600 15 ) 200 12 )

1400 ) ) weight normalized apparent 1200 12 150 9 (filled circles) and renal (open 1000

1 9 800 circles) clearance of bambuterol 100 6 Apparent clearance clearance Apparent Renal clearance 600 6 Renal clearance Apparent clearance (mL/min/kg body weight (mL/min/kg (mL/min/kg body weight (mL/min/kg (mL/min/kg body weight body (mL/min/kg 400 weight body (mL/min/kg and generated terbutaline 3 50 3 200 during oral administration of 0 0 0 0 0 20406080100 0 20406080100 bambuterol Weight (kg) Weight (kg)

Table 4. Normalized apparent (CLA/weight) and renal (CLR/weight) clearance of bambuterol and generated terbutaline during oral once-daily administration of bambuterol Variable Ethnic origin Age group Dose CLR/Weight CLA/Weight (mg) (mL/min/kg) (mL/min/kg) Bambuterol Caucasian Adults 10 3.28 398 20 2.59 307 30 2.17 191 Preschool 5 2.71 369 10 2.96 435 School 10 2.51 242 20 2.68 162 Oriental Preschool 2.5 . 739 5 . 631 School 10 3.08 253 20 3.13 145 Terbutaline Caucasian Adults 10 1.64 31.9 20 1.66 38.2 30 1.65 36.4 Preschool 5 3.34 82.7 10 3.30 98.8 School 10 2.61 65.4 20 2.7 69.4 Oriental Preschool 2.5 2.57 90.8 5 2.43 77.2 School 10 2.57 57.2 20 2.94 67.3

27 Mass balance and metabolism of 3H-formoterol in healthy men (Paper III)

The plasma peak of 3H-formoterol after intravenous (16.2 µg) and oral (37.8 µg) co- administration was attained within minutes, after which the concentration rapidly decreased in a poly-exponential manner. The metabolite peaks were lower and appeared later than the peak of formoterol (Figure 13).

Figure 13. Mean plasma concentration time curves of formoterol and its metabolites when inhalation of formoterol was mimicked. Linear (left) and logarithmic (right) y-axis scale. The horizontal line in the right graph indicates limit of quantification.

Mean total recovery was 86% of the administered formoterol dose, 62% in urine and 24% in faeces. Tritiated water was generated and because its in vivo turnover is slow, the terminal decline of total radioactivity was slow and dose recovery was incomplete during the sampling period. Formoterol was conjugated to inactive glucuronides and a previously unidentified sulphate. The phenol glucuronide of formoterol (FG1) was the main metabolite in urine (Figure 14).

Formoterol was also O-demethylated (Met1) and deformylated (Met2). Plasma exposure to these pharmacologically active metabolites was low. O-demethylated formoterol was seen mainly as inactive glucuronide conjugates and deformylated formoterol only as an inactive sulphate conjugate. Intact formoterol and O- demethylated formoterol dominated the recovery in faeces. Mean recovery of unidentified metabolites was 7.0% in urine and 2.0% in faeces.

28 Figure 14. Mean (columns) and individual (X) cumulative recovery in urine of formoterol and its metabolites.

Relative systemic dose potency and tolerability of inhaled formoterol and salbutamol in healthy subjects and asthmatics (Paper IV)

Systemically, formoterol was estimated to be 28-109 times as potent as salbutamol, depending on the measured variable; potency relations with respect to serum potassium suppression and the decrease in diastolic blood pressure are illustrated in Figure 15. The duration of systemic action seemed to differ marginally at approximately equieffective doses of formoterol and salbutamol. Systemic effects were well tolerated and tended to be more pronounced in the healthy subjects than in the asthmatic patients. Half-life of the pharmacologically active (R;R)-formoterol was longer than half-life of the inactive (S;S)-formoterol, 7.8 vs 5.5 h.

Assessment of a relative therapeutic index between inhaled formoterol and salbutamol in asthma patients (Paper V)

A linear approximation was considered sufficient to describe bronchodilation whereas a sigmoid approximation was more apt to describe serum potassium suppression. A bivariate non-linear mixed effect model based on these principles was fitted + simultaneously to all individual measurements of maximum FEV1 and minimum S-K in order to estimate the relative therapeutic index between formoterol and salbutamol. The fitted model is presented graphically superimposed on mean maximum values of FEV1 and mean minimum values of serum potassium concentration in Figure 16. The mean relative therapeutic index between formoterol given via Turbuhaler and salbutamol given via pMDI was estimated at 2.5. This ratio in favour of formoterol was not statistically significant, the 95% confidence interval for the estimate being 0.9–6.5.

29 Serum potassium Diastolic blood pressure

4.5 74.0

4.4

4.3

4.2 71.0

4.1

4.0 68.0 3.9

3.8

3.7 65.0 Healthy subjects Healthy 3.6

3.5

3.4 62.0 10 100 1000 10000 10 100 1000 10000 Dose (ug) Dose (ug)

4.5 74.0

4.4

4.3 71.0 4.2

4.1

4.0 68.0 3.9

3.8

3.7 65.0 3.6 Asthmatic patients 3.5

3.4 62.0 10 100 1000 10000 10 100 1000 10000 Dose (ug) Dose (ug) Figure 15. Mean serum potassium concentration and mean diastolic blood pressure 30 min after each subdivided dose fraction vs. the cumulative dose in healthy subjects and asthmatic patients. Dose fractions of formoterol were 13.5+13.5+27 µg (filled circles), and dose fractions of salbutamol were 300+300+600 µg (open squares), or 900+900+1800 (filled squares). Dotted lines indicate the dose of salbutamol shown to be equipotent with half the cumulative dose of formoterol. The solid line, parallel to the x-axis, indicates the 95% confidence interval for the estimate of equipotency.

25 4.1 O : Formoterol : Salbutamol 4.0 88 × 20 3.9

3.8 215 × 15 3.7

3.6 10 3.5 9 µg 1935 µg 9 µg 792 µg 5 3.4 1 10 100 1000 10000 1 10 100 1000 10000

Figure 16. Mean maximum FEV1 and minimum serum potassium concentration together with approximations of the dose response relationships. Dashed lines indicate the dose of salbutamol that is equipotent to formoterol 9 µg. The solid line, parallel to the x-axis, indicates the 95% confidence interval for that estimate. Note that relative increases in maximum FEV1 compared with placebo, not differences, are plotted against dose. The reason is that log transformed data were used in the statistical analysis.

30 Discussion

Clinical pharmacokinetics of bambuterol

Terminal elimination of orally administered bambuterol was shown to be slow and most likely absorption-rate limited (Paper I). This flip-flop behaviour of bambuterol was less clear during the first few hours after dosing (Figure 9) when a fraction was more rapidly absorbed than it was eliminated. Indeed, the model simultaneously fitted to literature data provided some evidence that absorption of bambuterol be multi-phasic; the predicted half-life of rapid absorption, using a biphasic absorption model, was only about 1.5 h (ln2/K81, Table 3), whereas the predicted half-life of slow absorption, about 14.5 h (ln2/K91), was in close agreement with the mean estimate of terminal half-life in Paper I. The slow absorption is probably a rate-limiting factor for elimination of the major bambuterol fraction.

The mean T½ of terbutaline generated from oral bambuterol was estimated at 22 h (Paper I). The mean after oral administration of plain terbutaline tablets as such is about 17 h (25). The model analysis indicated that elimination of terbutaline generated from bambuterol was slowed down after oral dosing of the prodrug, because biotransformation is sustained systemically and further more by first-pass mechanisms. The fitted model described well the terminal elimination rate of terbutaline after intravenous administration as such, but tended to overestimate that rate after administration of bambuterol. This is likely due to overestimation of terbutaline formation rate by the model, since the elimination rate of terbutaline derived from bambuterol seems to be governed by its formation rate.

A comparison of dosing-interval AUCs showed that terbutaline was in steady state after one week during treatment with bambuterol (Paper I). T½ of terbutaline as such was similar in Caucasian children and adults (55). Therefore, it is reasonable to assume that steady state for terbutaline is reached after one week of a once-daily bambuterol regimen in asthmatic children, too.

Dose proportionality was shown for terbutaline, i.e., the plasma exposure (AUC) of terbutaline was linearly correlated to the administered dose of bambuterol, but this was not the case for bambuterol (Papers I and II). Hydrolysis of bambuterol is to a great extent catalyzed by pseudocholinesterase. Bambuterol is a potent inhibitor of pseudocholinesterase and therefore partly inhibits its own metabolism (33); this is a conceivable reason for the observed pharmacokinetic non-linearity. The outcome of the modelling exercise suggested that about 60 % of systemically bioavailable terbutaline generated from a moderate oral dose of bambuterol was produced by slow presystemic biotransformation and that the rest was generated systemically. Considering the found dose linearity for terbutaline, any loss of presystemic capacity to generate terbutaline should have been compensated for systemically. Previous data in subjects with low pseudocholinesterase activity support this, since mean AUC0-24h of terbutaline after bambuterol 20 mg in that group was similar to the mean in subjects with normal pseudocholinesterase activity after the same dose (56).

31 It is often recommended that drug dosing to children, compared with adults, should be reduced in proportion to the children’s lower weight. This routine dose-adjusting principle in paediatric practice is mainly based on the assumption that the clearance of the drug be linearly proportional to body weight, whereas bioavailability is assumed to be independent of body weight. However, oral dosing of bambuterol based on body weight adjustment was shown not to be an effective strategy in children (Figure 12), primarily because weight-normalized clearance was higher in the children (Table 4). In fact, dosing on the basis of body weight could lead to undertreatment of children.

Clinical pharmacokinetics of formoterol

Systemic absorption of inhaled formoterol is rapid; the plasma peak due to substance overflow into the systemic circulation is seen within 15 min after dosing, after which the concentration rapidly declines (57). The plasma profile after simultaneously administered oral and intravenous dosing of formoterol, was shown to fairly well serve its purpose - to simulate inhalation (Paper III).

Glucuronidation of formoterol occurred mainly at the phenolic position, but also the benzylic OH position was conjugated (Paper III). This is similar to the adrenoceptor antagonist labetalol, where monoglucuronides were found at the benzylic or the phenolic positions (58, 59). In conformity with terbutaline (60), but in contrast to salbutamol (61), elimination of the pharmacologically active (R;R)- formoterol (62) was shown to be slower than elimination of the inactive (S;S)-formoterol (Paper IV). A similar trend, albeit not statistically significant, has been reported after administration of racemic formoterol via the dry powder inhaler Aerolizer (57). More rapid glucuronidation of (S;S)-formoterol could contribute to this enantioselectivity in the pharmacokinetics of formoterol (63). The capacity for glucuronidation is normally high and independent of age (64). Thus, if the capacity for phase-1 reactions, e.g. O- demethylation, were reduced, this would probably be metabolically compensated for by conjugation of formoterol.

Formoterol and O-demethylated formoterol, but no conjugates were shown to be excreted in faeces (Paper III). Deconjugation in the intestine of systemically formed and excreted conjugates (65) might explain this difference in proportions of formoterol and metabolites in faeces compared with urine and plasma.

Despite a long duration of bronchodilation, the systemic duration of action of inhaled formoterol, represented by serum potassium suppression, was not prolonged compared with an inhaled systemically equieffective salbutamol dose. Thus, the systemic exposure to inhaled therapeutically relevant doses of formoterol, and to some extent to pharmacologically active metabolites, seems to be too low to prolong the systemic action compared with a short-acting analogue.

32 The mean relative therapeutic index between formoterol (Oxis) 4.5–54 µg given via Turbuhaler and salbutamol 200–1800 µg given via pMDI was estimated at 2.5 (Paper V). Thus, systemically mediated side-effects of inhaled formoterol may be less pronounced than with an equieffective bronchodilating dose of inhaled salbutamol. The octanol/water partition coefficient of formoterol is approximately 200 times that of salbutamol (66). Although not being the only factor behind the observation, this physicochemical difference is likely to contribute to the observed difference in duration of bronchodilation between these two drugs after administration via the lung (2). However, because the measured ratio did not statistically significantly differ from unity, further studies of this issue are desirable.

Concluding remarks

Regularity is a key element of maintenance drug treatment; either the goal is to exert continuous or intermittent effects. Compliance is crucial for treatment success, particularly for lengthy treatment of chronic diseases. Education plays a prominent role in the patient’s disease awareness, comprehension and ultimately compliance with a prescribed drug regimen. The importance of simplicity should therefore not be underestimated. Once- or twice-daily intake of a drug is always easier to comply with than regimens requiring more frequent dosing. Regular treatments of asthma and chronic obstructive pulmonary disease with long-acting β2-agonists are examples where continuously exerted effect can be a desirable outcome (67, 68). This thesis illustrates how that goal can be achieved by two different approaches: oral administration of the prodrug bambuterol and inhalation of formoterol. The pharmacological challenge was to prolong duration of bronchodilation without getting troublesome side-effects. To achieve this, the systemic plasma concentration of terbutaline, generated from bambuterol, and local lung concentration of formoterol are to be kept within therapeutically relevant ranges for extended periods of time. The systemic plasma concentration of generated terbutaline was used as a marker of the sustained bronchodilating potential of bambuterol, whereas systemic and local effect variables were used to illustrate the selective long-acting potential of inhaled formoterol.

Recommended doses of bambuterol, 10-20 mg once daily in adults, produced plasma concentrations of the active moiety terbutaline, normally being within therapeutically relevant limits. Achieving this by once-daily dosing is a unique property of bambuterol among prodrugs of β2-agonists. Dose proportionality with respect to generated terbutaline makes dosing of bambuterol predictable. Children should be given higher doses than indicated by their lower body weight compared with adults. Parametric and non-parametric pharmacokinetic analysis of plasma concentrations showed that absorption of bambuterol is slow and multi-phasic and indicated that biotransformation to terbutaline, which is also slow, occurs both presystemically and systemically. The steady-state level of terbutaline was higher than after a single-dose, indicating that the full effect of bambuterol cannot be expected until after a few days.

33 Although bronchodilation is sustained by inhaled formoterol, Paper IV in this investigation suggested that the duration of systemically-mediated cardiovascular or metabolic side-effects of formoterol do not differ from inhaled salbutamol. There was a trend in Paper V that the magnitude of systemic side-effects may even be less pronounced after inhalation of formoterol compared with a locally equieffective dose of inhaled salbutamol. The local action of inhaled formoterol is probably maintained by substance retained in the lungs. Once absorbed into the systemic circulation, formoterol is rapidly eliminated or transformed into inactive metabolites - primarily glucuronide conjugates. This could explain why systemically-mediated side-effects were less pronounced and wore off more rapidly than the bronchodilating effect of inhaled formoterol.

Both approaches to sustaining stimulation of β2-adrenoceptors have their pros and cons. Bambuterol can be dosed orally once daily, but full effect is reached slowly. The effect of formoterol is reached within a few minutes, but administration must occur via the lungs, often twice daily. Both treatments, however, give 24-h symptom relief during regular treatment.

34 ACKNOWLEDGEMENTS

This dissertation has been made possible by the kind help and generous support from my colleagues at AstraZeneca R&D Lund. Particularly, I wish to express my gratitude to

Lars Nyberg, my mentor, tutor, and friend, for introducing me to pharmacokinetics, many inspiring discussions, co-authorship, and inexhaustible support over the years.

Lars Borgström, my understanding boss, mentor in the field of inhalation pharmacokinetics, daily encouraging discussion partner, and patient reader of all manuscripts.

Professor Margareta Hammarlund-Udenaes, my supervisor, who encouraged me to complete this work and gave valuable pieces of advice.

Professor Lennart Paalzow, who was supportive during the initial years.

My co-authors, particularly Per Larsson, for his sound and realistic attitude to clinical research, and for his biomathematical skills and support.

Bertil Waldeck and Nils Grundström, for their pharmacodynamic skills and inspiring discussions.

Bengt Johansson, for his linguistic suggestions and improvement of clarity of the text.

Cristina Mardell and Vivian Kristofferson, for their excellent study co-ordination.

Anna Malmgren, for her support during the final preparation of the thesis.

The personnel at Data Management, the Clinical Phase I Unit, Pharmaceutical & Analytical R&D, and Development Drug Metabolism & Bioanalysis.

AstraZeneca, for its benevolent attitude to my dissertational work.

I also wish to thank my parents.

And finally, I thank my wife Agnes.

35 REFERENCES

1. Pauwels R, Löfdahl C, Postma D, et al. Effect of inhaled formoterol and budesonide on exacerbations of asthma. N Engl J Med 1997;337:1405-1411. 2. Anderson G, Lindén A, Rabe K. Why are long acting beta-adrenoceptor agonists long acting? Eur Respir J 1994;7:569-78. 3. Lands A, Arnold A, McAuliff J, Luduena F, Brown TJ. Differentiation of receptor systems activated by sympathomimetic amines. Nature 1967;214:597-598. 4. Bergman J, Persson H, TI WK. Two new groups of selective stimulants of adrenergic beta-receptors. Experientia 1969;25:899-901. 5. Hartley D, Jack D, Lunts L, Ritchie A. New class of selective stimulants of beta- adrenergic receptors. Nature 1968;219:861-862. 6. Larsson S, Svedmyr N, Thiringer G. Lack of bronchial beta-adrenoceptor resistance in asthmatics during long-term treatment with terbutaline. J Allergy Clin Immunol 1977;59:93-100. 7. Barnes P. Beta-adrenergic receptors and their regulation. Am J Respir Crit Care Med 1995;152:839-60. 8. Bowman W. Effects of adrenergic activators and inhibitors on the skeletal muscles. In: Szekeres L, ed. Adrenergic activators and inhibitors. Handbook of Experimental Pharmacology. Berlin: Springer-Verlag, 1980:47-128. vol 54/2). 9. Gennari F. Hypokalemia. N Engl J Med 1998;339:451-58. 10. Genuth S. The adrenal glands. In: Berne R, Levy M, eds. The Endocrine system. St Louis: Mosby year book, 1993:949-979. 11. Guyton A. The autonomic nervous system; the adrenal medulla. In: Textbook of Medical Physiology 8th ed. Philadelphia: W.B. Saunders Company, 1991: 667-678. 12. Andreoli T. Disorders of fluid volume, electrolytes, and acid-base balance. In: Wyngaarden J, Smith L, eds. Textbook of Medicine. Philadelphia: W.B. Saunders Company, 1988:528-558. 13. Lipworth B. Risks versus benefits of inhaled beta2-agonists in the management of asthma. Drug Saf 1992;7:54-70. 14. Newnham D, Grove A, McDevitt D, Lipworth B. Subsensitivity of bronchodilator and systemic beta-2 adrenoceptor responses after regular twice daily treatment with eformoterol dry powder in asthmatic patients. Thorax 1995;50:497-504. 15. Tötterman K, Huhti L, Sutinen E, et al. Tolerability to high doses of formoterol and terbutaline via Turbuhaler for 3 days in stable asthmatic patients. Eur Respir J 1998;12:573-579. 16. Lamont H, Van der Straeten M, Pauwels R, Moerman E, Bogaert M. The combined effect of theophylline and terbutaline in patients with chronic obstructive airway diseases. Eur J Respir Dis 1982;63:13-22. 17. Billing B, Dahlqvist R, Garle M, Hörnblad Y, Ripe E. Separate and combined use of terbutaline and theophylline in asthmatics. Effects related to plasma levels. Eur J Respir Dis 1982;63:399-409. 18. Van den Berg W, Leferink J, Tabingh-Suermondt W, et al. Terbutaline serum concentrations related to different lung function parameters and beta-receptor function. Int J Clin Pharmacol Ther Toxicol 1983;21:24-30. 19. Fuglsang G, Pedersen S, Borgström L. Dose-response relationships of intravenously administered terbutaline in children with asthma. J Pediatr 1989;114:315-320.

36 20. Rowland M, Tozer T. Pharmacologic Response. Clinical Pharmacokinetics, Concepts and Applications. Third ed. Philadephia: Lea & Febiger, 1995:340-366. 21. Levy G. Kinetics of pharmacological effects. Clin Pharmacol Ther 1966;7:366-372. 22. Coleman R, Johnson M, Nials A, Vardey C. Exosites: their current status, and their relevance to the duration of action of long-acting B2-adrenoceptor agonists. Trends Pharmacol Sci 1996;17:324-330. 23. Tasaka K. Formoterol (Atock): a new orally active and selective beta2-receptor stimulant. Drugs Today 1986;22:505-519. 24. Löfdahl C, Svedmyr N. Formoterol fumarate, a new beta2-adrenoceptor agonist. Acute studies of selectivity and duration of effect after inhaled and oral administration. Allergy 1989;44:264-271. 25. Nyberg L. Pharmacokinetic parameters of terbutaline in healthy man. An overview. Eur J Respir Dis 1984;65 (suppl. 134):149-160. 26. Svensson L-Å. Drug Latentiation. Studies of catechol esters of certain diacids . Lund: Lund, 1971. 24 p. 27. Kristofferson J, Svensson L, Tegnér K. Drug latentiation of terbutaline. Acta Pharm Suec 1974;11:427-38. 28. Hörnblad Y, Ripe E, Magnusson PO, Tegnér K. The metabolism and clinical activity of terbutaline and its prodrug ibuterol. Eur J Clin Pharmacol 1976;10:9-18. 29. Olsson O, Svensson L. New lipophilic terbutaline ester prodrugs with long effect duration. Pharm Res 1984;3:19-23. 30. Tunek A, Levin E, Svensson L. Hydrolysis of 3H-bambuterol, a carbamate prodrug of terbutaline, in blood from humans and laboratory animals in vitro. Biochem Pharmacol 1988;37:3867-3876. 31. Svensson L, Tunek A. The design and bioactivation of presystemically stable prodrugs. Drug Metab Rev 1988;19:165-194. 32. Lindberg C, Roos C, Tunek A, Svensson L. Metabolism of bambuterol in rat liver microsomes: identification of hydroxylated and demethylated products by liquid chromatography mass spectrometry. Drug Metab Dispos 1989;17:311-322. 33. Tunek A, Svensson L. Bambuterol, a carbamate ester prodrug of terbutaline, as inhibitor of cholinesterase in human blood. Drug Metab Dispos 1988;16:759-64. 34. Nyberg L, Rosenborg J, Weibull E, Nilsson M, Jönsson S, Kennedy B-M. Pharmacokinetics of bambuterol in healthy subjects. Br J Clin Pharmacol 1998;45:471-478. 35. Clemmensen I, Pedersen B, Ravn S, Laursen L, Weeke B. Bambuterol: clinical effects of three doses of bambuterol once daily in asthmatic patients. Allergy 1988;43:573-6. 36. Vilsvik J, Langaker O, Persson G, Ringdal N, Kvelstad G, Schaanning J. Bambuterol: a new long acting bronchodilating prodrug. Ann Allergy 1991;66:315-9. 37. Holstein-Rathlou N, Laursen L, Madsen F, Svendsen U, Gnosspelius Y, Weeke B. Bambuterol: Dose response study of a new terbutaline prodrug in asthma. Eur J Clin Pharmacol 1986;30:7-11. 38. Petrie G, Chookang J, Hassan W, et al. Bambuterol: effective in nocturnal asthma. Respir Med 1993;87:581-585.

37 39. D'Alonzo G, Smolensky M, Feldman S, Gnosspelius Y, Karlsson K. Bambuterol in the treatment of asthma. A placebo- controlled comparison of once-daily morning vs evening administration. Chest 1995;107:406-412. 40. Persson G, Baas A, Knight A, Larsen B, Ohlsson H. One month treatment with the once daily oral β2-agonist bambuterol in asthmatic patients. Eur Respir J 1995;8:34-39. 41. Fugleholm A, Ibsen T, Laxmyr L, Svendsen U. Therapeutic equivalence between bambuterol, 10 mg once daily, and terbutaline controlled release, 5 mg twice daily, in mild to moderate asthma. Eur Respir J 1993;6:1474-1478. 42. Persson G, Pahlm O, Gnosspelius Y. Oral bambuterol versus terbutaline in patients with asthma. Curr Ther Res Clin Exp 1995;56:457-465. 43. Larsén K, Schmeckel B. Tremor in healthy volunteers after bambutreol and terbutaline CR-tablets. Eur J Clin Pharmacol 1993;45:303-305. 44. Gunn SD, Ayres JG, McConchie SM. Comparison of the efficacy, tolerability and patient acceptability of once-daily bambuterol tablets against twice-daily controlled release salbutamol in nocturnal asthma. Eur J Clin Pharmacol 1995;48:23-28. 45. Crompton G, Ayres J, Basran G, et al. Comparison of oral bambuterol and inhaled salmeterol in patients with symptomatic asthma and using inhaled corticosteroids. Am J Respir Crit Care Med 1999;159:824-28. 46. Wallaert B, Brun P, Ostinelli J, et al. A comparison of two long-acting ß-agonists, oral bambuterol and inhaled salmeterol, in the treatment of moderate to severe asthmatic patients with nocturnal symptoms. Respir Med 1999;93:33-38. 47. Brittain R, Dean C, Jack D. Sympathomimetic bronchdilator drugs. Pharmac Ther B 1976;2:423-462. 48. Palmqvist M, Persson G, Lazer L, Rosenborg J, Larsson P, Lötvall J. Inhaled dry- powder formoterol and salmeterol in asthmatic patients: Onset of action, duration of effect and potency. Eur Respir J 1997;10:2484-2489. 49. Guhan A, Cooper S, Oborne J, Lewis S, Bennet J, Tattersfield A. Systemic effects of formoterol and salmeterol: a dose response comparison in healthy subjects. Thorax 2000;55:650-656. 50. Lindberg C, Jönsson S, Paulson J. Determination of bambuterol, a prodrug of terbutaline in plasma and urine by gas chromatography mass spectrometry. Biomed Environ Mass Spectrom 1990;19:218-224. 51. Kennedy B, Blomgren A, Edholm L, Roos C. Quantitative determination of terbutaline in human plasma after administration of bambuterol using coupled columns and electrochemical detection. Chromatographia 1987;24:895-899. 52. Beal S, Sheiner L. (Eds.). NONMEM users guides. NONMEM project group. University of California, San Francisco, CA, 1992. 53. Gardner R, Hankinson J, Clausen J, Crapo R, Johnson R. Standardization of spirometry- 1987 update. Am Rev Respir Dis 1987;136:1285-1298. 54. Fieller E. Some problems in biological assay. J R Statist Soc B 1954;16:175-185. 55. Hultquist C, Lindberg C, Nyberg L, Kjellman B, Wettrell G. Pharmacokinetics of intravenous terbutaline in asthmatic children. Dev Pharmacol Ther 1989;13:11-20. 56. Bang U, Nyberg L, Rosenborg J, Viby-Mogensen J. Pharmacokinetics of bambuterol in subjects homozygous for the atypical gene for plasma cholinesterase. Br J Clin Pharmacol 1998;45:479-484. 38 57. Lecaillon J, Kaiser G, Palmisano M, Morgan J, Della Cioppa G. Pharmacokinetics and tolerability of formoterol in healthy volunteers after single high dose of Foradil dry powder inhalation via Aerolizer. Eur J Clin Pharmacol 1999;55:131-138. 58. Alton K, Chan T, Pramanik B. Urinary metabolites of (R),(R)-labetalol. Drug Metab Dispos 1994;22:866. 59. Niemeijer N, Gerding T, De Zeeuw R. Glucuronidation of labetalol at the hydroxy positions by bovine liver microsomes. Isolation, purification, and structure elucidation of the glucuronides of labetalol. Drug Metab Dispos 1991;19:20. 60. Borgström L, Nyberg L, Jönsson S, Lindberg C, Paulsson J. Pharmacokinetic evaluation in man of terbutaline given as separate enantiomers and as the racemate. Br J Clin Pharmacol 1989;27:49-56. 61. Ward J, Dow J, Dallow N, Eynott P, Milleri S, Ventresca P. Enantiomeric disposition of inhaled intravenous and oral racemic salbutamol in man - no evidence of enantiomeric lung metabolism. Br J Clin Pharmacol 2000;49:15-22. 62. Trofast J, Österberg K, Källström B-L, Waldeck B. Steric aspects of agonism and antagonism at ß-adrenoceptors: Synthesis of and pharmacological experiments with the enantiomer of formoterol and their diastereomers. Chirality 1991;3:443-50. 63. Zhang M, Fawcett J, Kennedy J, Shaw J. Stereoselective glucuronidation of formoterol by human liver microsomes. Br J CLin Pharmacol 2000;49:152-7. 64. Kroemer H, Klotz U. Glucuronidation of drugs. Clin Pharmacokinet 1992;23:292-310. 65. Gibson G, Skett P. Pharmacological and toxicological aspects of drug metabolism. In: Gibson G, Skett P, eds. Introduction to drug metabolism. London: Blackie Academic and Professional, 1994:157-179. 66. Ochsner M, Jaekel K, Mutz M, Anderson G, Johan E. Comparative biophysical analysis of the interaction of bronchodilating beta2-adrenoceptor agonists with lipid membranes. Eur J Med Chem 1999;34:451-462. 67. Anonymous. Global Strategy for Asthma Management and Prevention. NHLBI/WHO Workshop Report (1998 revision). Global Initiative for Asthma. National Institutes of Health, 1998:vol 95-3659). 68. Pauwels R, Buist A, Calverley P, Jenkins C, Hurd S. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease - NHLBI/WHO Global initiative for chronic obstructive pulmonary disease (GOLD) workshop summary. Am J Respir Crit Care Med 2001;163:1256-1276.

39 APPENDIX

Program code used to model the pharmacokinetics of bambuterol and terbutaline:

$PROBLEM Bambuterol (iv+po) & terbutaline (iv): C-B, C-T, C-T-B $INPUT ID AMT TIME RATE EVID CMT ADDL II DV FLAG VAR TYPE CP MIN $DATA data_all.prn IGNORE=# $SUBROUTINES ADVAN5 $MODEL COMP=(DEFOBS1) COMP=(PERIPH1) COMP=(PERIPH2) COMP=(DEFOBS2) COMP=(PERIPH3) COMP=(PERIPH4) COMP=(PERIPH5) COMP=(DEPOT1) COMP=(DEPOT2) COMP=(DEPOT3) $PK X3=0 X4=0 IF(VAR.EQ.1)X3=1 IF(VAR.EQ.2)X4=1 ;------;BASIC PK OF BAMBUTEROL ;------K10=THETA(1)*(1+ETA(1)) K12=THETA(2)*(1+ETA(2)) K21=THETA(3)*(1+ETA(3)) K13=THETA(4)*(1+ETA(4)) K31=THETA(5)*(1+ETA(5)) ;------;BASIC PK OF TERBUTALINE ;------K40=THETA(6)*(1+ETA(6)) K45=THETA(7)*(1+ETA(7)) K54=THETA(8)*(1+ETA(8)) K46=THETA(9)*(1+ETA(9)) K64=THETA(10)*(1+ETA(10)) ;------;VOLUMES OF DISTRIBUTION ;------S1 =THETA(11)*(1+ETA(11)) S4 =THETA(11)*(1+ETA(11)) ;------;SYSTEMIC GENERATION OF TERBUTALINE FROM BAMBUTEROL ;------K74=THETA(12)*(1+ETA(12)) K17=THETA(13)*(1+ETA(13)) ;------;GI ABSORPTION OF BAMBUTEROL (K81, K91) AND ;PRESYSTEMIC RATE OF TERBUTALINE GENERATION (K104) ;------K81 =THETA(14)*(1+ETA(14)) K91 =THETA(15)*(1+ETA(15)) K104=THETA(15)*(1+ETA(15)) F8 =THETA(16)*(1+ETA(16)) F9 =THETA(17)*(1+ETA(17)) F10 =THETA(18)*(1+ETA(18)) $ERROR IF (EVID.EQ.0.AND.F.GT.0) THEN IPRED=LOG(F) ELSE IPRED=LOG(F+1) 40 ENDIF IRES=DV-IPRED IWRES=IRES Y=IPRED+X3*EPS(1)+X4*EPS(2) ; K10:1 K12:2 K21:3 K13:4 K31:5 $THETA (0, 5.36) (0, 5.29) (0, 1.79) (0, 3.23) (0, 0.356) ; K40:6 K45:7 K54:8 K46:9 K64:10 $THETA (0, 1.11) (0, 0.635) (0, 0.065) (0, 1.79) (0, 0.991) ; S1,4:11 $THETA (0, 9.82) ; K74:12 K17:13 $THETA (0, 0.270) (0, 3.18) ; K81:14 K91,104:15 F8:16 F9:17 F10:18 $THETA (0, 0.479) (0, 0.0533) (0, 0.034) (0, 0.053) (0, 0.058) $OMEGA 0.1539 ;ETA(1) $OMEGA 0 FIXED ;ETA(2) $OMEGA 0 FIXED ;ETA(3) $OMEGA 0 FIXED ;ETA(4) $OMEGA 0 FIXED ;ETA(5) $OMEGA 0.1044 ;ETA(6) $OMEGA 0 FIX ;ETA(7) $OMEGA 0 FIX ;ETA(8) $OMEGA 0 FIX ;ETA(9) $OMEGA 0 FIX ;ETA(10) $OMEGA 0.2 ;ETA(11) $OMEGA 0.1780 ;ETA(12) $OMEGA 0.1539 ;ETA(13) $OMEGA 0 FIX ;ETA(14) $OMEGA 0 FIX ;ETA(15) $OMEGA 0.2 ;ETA(16) $OMEGA 0.2 ;ETA(17) $OMEGA 0.2 ;ETA(18) $SIGMA 0.2112 0.1507 $EST NOABORT POSTHOC MAXEVAL=9999 PRINT=5 $COV

Basic goodness of fit plots for run 3_cov

357357 2 2 848216126 84841613573567 773535 4 76721352 5347164821862 412483586 4 3547816 2 4 3516487531672 7355468716 2 2653164872 71553753573363537537236841 2 1127515371636236751237348125 37353838737573531745175863786125487164416246812486122 71723146127562356834527853624537432542861483 3 7 5855753825746782468317626731617276182614612263421232 53574356732846713864624781361726587345613473561748575313 2 7 713752867533768365784531421347852347865432861672412514261628843414333 3 2 18263365835683472315842615726348752347864368218761252316741682784 584345345775112758784652756843815465731845216326148261424621348 8 45132178758462847561836781542624657138563468123754458 5 31743513775375686842748761438652161261421 7 1531354632756852761824751458754386514716 35784845883841456267212112 56483428845641734715736577717 7 783535673175836212 6412 5 661374527324682741253856 5 37 857836837375164351641224724 6 6768678534514612736458126 0 85576684642582838147122614 726 0 7 57648452716823878321717535261 7 78446112176 3885414846543831317 8 1874 76 876 6 6 8 7 375 7 IPRE 1 26 PRED 7 -2 184 26 -2 3584 6 7 2 6 7 2 -4 4 8 -4 4 8 6 -6 8 6 -6 8

-6 -4 -2 0 2 4 -6 -4 -2 0 2 4 DV DV

1.2 5 4 6 1.0 7 1 533616 6 61 7 7 2 58 7 2 6 8 1 8 4 4 1 1 32161 7 7 7 5 54 0.8 287418782 6176 7 65 353 5 4 5 86265312421483216112 732 7 83 4 5353 4 2 638475218368741864242187652 52 31 8 6 8324864 54 54 4 | IWRE | | IWRE 7564217538164875164723276515436411 42437273278654272 2 538765342187654317286543876427275315452764287832 632 428628434 4 5 1 21 7 0.6 0 164358643728765321872876541643287541684324842 62638 32 7426537514387841537653274743287638584351 2 57657672642 532 861 2 1 2531654234364731742131685434 75 1 87651553 6 7863 7862 71 835235328648638726 352 4 4 8 7185111785 86321 63 6 1 WRES 41835546 75 55 8 61 7 71 1 0.4 642585 87 8 6 7 3 1 1 8 53535 7561 6 4 61 1 3 -2 75 4 5 7 4 7 63 0.2 8 7 -4 0.0 7 -6 -4 -2 0 2 4 0 50 100 150 IPRE TIME

Figure 17. Pharmacokinetic modelling of plasma bambuterol and terbutaline, basic goodness of fit plots. DV=measured plasma concentration; PRED=predicted concentration based on population parameter mean estimates; IPRE=predicted concentration based on individual parameter mean estimates; WRES=weighted residual; IWRE=individual weighted residual

41