Can lung deposition data act as a surrogate for the clinical response to inhaled asthma drugs?

Abstract

Studies involving the direct measurement of clinical response to inhaled asthma drugs, especially inhaled corticosteroids, may be very difficult to conduct. However, the deposition of drug in the lungs may be considered as a measure of local bioavailability, and may be quantified by radionuclide imaging techniques, or for some drugs by pharmacokinetic methods. This paper reviews evidence for considering lung deposition data as a surrogate for the clinical response to inhaled asthma drugs, based mainly upon a series of case histories. The appropriate use of lung deposition data in regulatory packages, especially to document the equivalence or comparability of two products, offers the possibility of significant time saving in the drug development process, and hence a faster drug development programme for inhaled asthma products.

Development of new asthma products

Many new inhaled drug products are being developed for the treatment and prevention of asthma attacks. Since the mid-1950s inhaled asthma therapy has relied chiefly upon the pressurized metered dose inhaler (pMDI) powered by chlorofluorocarbon (CFC) propellants. However, the banning of CFCs under the Montreal protocol, together with both the recognition of other limitations of the pMDI and the move towards generic inhaler products, has led to much activity within the pharmaceutical industry in a rush to develop and market alternative inhalers [1].

Inhaled asthma products can be classified essentially into those which relieve acute bronchospasm (mainly the β-adrenoceptor agonist bronchodilators such as salbutamol, terbutaline and salmeterol) and those which prevent asthma attacks (including the inhaled corticosteroid drugs such as beclomethasone dipropionate (BDP), budesonide and fluticasone propionate, and the cromones sodium cromoglycate and nedocromil sodium). One way of showing the effectiveness of a new inhaler product is to measure clinical response directly, but this apparently simple objective is fraught with problems. In the case of bronchodilators, a rapid therapeutic response may be measured as an improvement in spirometric tests of lung function such as forced expiratory volume in one second (FEV₁) and peak expiratory flow (PEF). However, it is very easy to reach the top of a dose–response curve for normal treatment doses of inhaled bronchodilators, so that spirometric tests may fail to detect important differences in drug delivery between two inhaled products [2].

In the case of inhaled corticosteroids, no rapid response occurs, and the usual approach for comparing two inhalers is to conduct trials of at least 4 weeks' duration, with the patients allocated into parallel groups. It has been suggested that there should be a lengthy screening period before the trial when the lowest maintenance dose is defined for individual patients in order to ensure that they will respond to the treatment in the trial itself [3]. Steroid-naive patients have been proposed as the most suitable study population, but such patients are now difficult to find in western Europe. Not only is a clinical trial of this type very complex and unwieldy, but it also lacks incisiveness; the variability of response to inhaled corticosteroids is high, so that it may be necessary to recruit populations in excess of 100 patients to each group [4]. Further, the parameters measured in such a trial, primarily consisting of diary card recordings of PEF, symptom scores and use of bronchodilator medication at home, are relatively imprecise clinical endpoints.

For these reasons it is desirable to investigate the role of other types of study that can assess the similarity or difference between two inhaled asthma products (e.g. innovator and generic pMDIs, or pMDI vs dry powder inhaler), so that the burden of documentation required to register new inhaler devices or formulations can be simplified. However, while European guidelines [5] allow for the use of ‘local bioavailability’ studies for topically acting drugs, the measure chosen must be validated. Lung deposition may be considered as a measure of local bioavailability for inhaled asthma drugs which act on the airway surface itself. This article reviews the evidence for use of lung deposition data as a surrogate for the clinical response to inhaled asthma drugs, based in particular upon a series of case histories.

Measuring lung deposition of inhaled asthma drugs

Methods available for assessing lung deposition

Methodologies for quantifying the deposition of drugs from inhaler devices are listed in Table 1, and were recently reviewed by the British Association for Lung Research (BALR) in a Consensus Statement [6]. In vitro tests performed primarily by impactors such as the Andersen sampler quantify the ‘fine particle mass’ or ‘fine particle dose’, i.e. drug contained in particles or droplets smaller than about 5 μm diameter, and which are theoretically small enough to enter the lungs. However, in vitro assessments should be considered mainly as quality control measures, as they are known to overestimate systematically the amount of drug reaching the lungs in vivo [7]. Direct measures of drug in lung tissue are generally not possible in man, while recovery of drug in bronchoalveolar lavage fluid is an unpleasant and invasive procedure. Assessments of the difference between the amounts of drug in inspired and expired air [8], or the amount of drug collected on a filter placed between the inhaler and the mouth [9] yield measures of total deposition in the body, but cannot distinguish between fractions deposited in the lungs and in the oropharynx. Pharmacokinetic methods and imaging techniques therefore offer the best prospects for accurately quantifying lung deposition in vivo.

Table 1.

Methods for estimating lung deposition.

Pharmacokinetic methods

Figure 1 shows a schematic representation of the fate of an inhaled dosage form in man. Systemic drug levels will in general result from absorption of drug via the lungs and via the gastrointestinal (g.i.) tract, and systemic drug levels may not reflect lung deposition of drug unless g.i. absorption can be excluded. Nevertheless, several pharmacokinetic methods have been described for assessing drug deposition in the lungs, based on the quantification of drug in plasma or urine. For some drugs, such as sodium cromoglycate, oral bioavailability is negligible, and plasma levels can be used as a comparative measure of the relative lung depositions between treatment regimens [11]. Using the ‘charcoal block’ method [12], the amount of drug deposited in the lungs can be quantified for a limited range of drugs including terbutaline sulphate and budesonide following the blocking of g.i. absorption by a swallowed charcoal slurry. Data similar to those obtained with the charcoal block method can be obtained for drugs whose oral bioavailability is accurately known [13]. The concentrations of drug in either the plasma [14] or in the urine [15] during the first 30 or 60 min after inhalation may also constitute an index of lung deposition for some drugs, since the contribution of swallowed drug to systemic levels is negligible during these time periods. The major limitations of pharmacokinetic methods are that they are highly drug specific, and that they do not provide any data about the pattern of regional deposition within the lungs themselves.

Figure 1.

Schematic diagram showing the fate of a drug dose given by inhalation. A dose from an inhaler may be deposited in either the lungs or in the oropharynx, and hence may find its way into the systemic circulation following absorption from either the lungs or the gastrointestinal tract. From Borgström [10], with permission.

Imaging techniques

The radionuclide imaging methods of gamma scintigraphy, single photon emission computed tomography (SPECT) and positron emission tomography (PET) enable whole lung deposition to be quantified and regional lung deposition to be assessed. The amount of drug deposited in the lungs from inhalers, and probably its distribution within the lungs themselves, are critical to the success of asthma therapy, since these parameters will determine directly the clinical efficacy of drugs given by inhalation. Gamma scintigraphy (planar lung imaging) is the most practical and widespread of these methods, and can accurately quantify the amount of drug delivered to the lungs for any inhaled drug product [16–18]. Gamma scintigraphy generally uses formulations radiolabelled with the radionuclide ^99mTc, which results in very low radiation doses as a result of its decay characteristics (6 h half-life, and mainly gamma rays with few beta rays emitted in the decay process). Following inhalation, the lungs and oropharynx are imaged in two dimensions by a large radiation detector (gamma camera) connected to a data processing system in which images are stored as a matrix of picture elements (pixels). Prior to the study itself, in vitro radiolabelling validation experiments are conducted to prove that the ^99mTc radiolabel is a precise marker for the drug across the full range of particle size bands. Regional lung deposition patterns are assessed by dividing the lung fields into a series of zones representing primarily deposition in large, medium and small diameter airways. Typical scintigraphic images comparing depositions from a pMDI and a pMDI coupled to a Nebuhaler large volume spacer device are shown in Figure 2.

Figure 2.

Scintigraphic images comparing lung and oropharyngeal depositions for pMDI and pMDI plus Nebuhaler spacer device. From Newman & Newhouse [19], with permission.

The three-dimensional imaging techniques of SPECT and PET offer the potential advantage of being able to relate deposition patterns within the lungs more precisely to airways of different sizes than is possible in gamma scintigraphy [20]. Recent studies using SPECT imaging have attempted to link the regional patterns of aerosol distribution with deposition in specific airway generations [21]. PET imaging has a further potential benefit, in that it is possible to chemically incorporate a radiolabel directly into the structure of the drug molecule [22]. Both SPECT and PET are powerful research tools, but both are still in development for the assessment of drug delivery to the lungs, and the practical problems associated with applying these methods to multidose inhaler devices have recently been reviewed [20]. These problems relate primarily to the need to use large amounts of radiotracer and/or long scanning times for SPECT, and the difficulties of working with very short lived cyclotron-produced radionuclides for PET.

Consequently, the BALR Consensus Statement [6] concluded that while SPECT and PET have promise for the future, gamma scintigraphy is the imaging modality of choice for assessing drug delivery to the lungs at the present time.

Relationship between drug deposition and response

The deposition of asthma drugs in the lungs should predict their beneficial clinical effects, since these drugs act locally on the airway surface to treat or prevent an asthmatic attack, and the assessment of lung deposition quantifies the extent to which drug is targeted towards the required site of action. Further, we would expect deposition data to be predictive of side-effects of inhaled drugs. The local side-effects of inhaled corticosteroids (oral candidiasis and dysphonia) are dependent upon the amount of drug deposited in the oropharynx, while the systemic side-effects will be proportional to total body exposure, which will depend in turn upon the sum of lung deposition plus the product of oropharyngeal deposition and oral bioavailability.

The relationship between lung deposition of inhaled asthma drugs and their clinical effects has been the subject of two recent major review articles [3, 23]. Selroos et al. [3] showed that devices with superior lung deposition also give a superior clinical effect, and stated that ‘these differences in deposition figures have been reflected in the results of most single-dose cross-over studies with bronchodilator substances’. Pauwels et al. [23] concluded that ‘there is in general a good relationship between pulmonary deposition and pulmonary effects for both bronchodilators and steroids, although the documentation is still rather limited and more extensive for bronchodilators’.

In making comparisons between lung deposition and clinical response, several problems are apparent. Many studies fail to show a correlation between deposition of drug in the lungs and clinical response because they have been conducted with drug doses which approach the top of the dose-response curve. Ideally, clinical comparisons should involve more than one dose level, in order to ensure that at least the lower doses will produce responses on the steep portion of the curve. In addition, the actual site within the lungs at which inhaled asthma drugs act is poorly defined. Receptors for β-adrenoceptor agonists appear to be located throughout the airways, and total lung dose seems to be more important than the regional deposition of drug [24]. However, muscarinic receptors are located primarily in larger central airways, and it may be necessary for antimuscarinic drugs to be delivered preferentially to these airways. The inflammatory process in asthma is probably present throughout the bronchial tree, and hence anti-inflammatory drugs such as inhaled corticosteroids probably need to be delivered throughout the airways, including the bronchioles [25].

The link between deposition of drug and clinical response may not have been detected for some products, since the only way of examining this link was to compare lung deposition data obtained at one study centre with clinical response data obtained at another study centre, and possibly in a different subject group. For instance, a scintigraphic study carried out on the Spacehaler (Medeva, formerly known as Gentlehaler) showed similar whole lung and regional lung depositions for 100 μg salbutamol delivered to a group of asthmatics by conventional pMDI actuator and by Spacehaler, together with identical bronchodilatation [26]. However, a pharmacokinetic study conducted in healthy volunteers at another centre showed higher plasma levels of salbutamol during the first hour after inhalation for the Spacehaler, together with a greater degree of finger tremor and larger changes in serum potassium [14]. These data could be interpreted as showing that lung deposition data do not predict pharmacodynamic effects, but it would only be justified to reach this conclusion as the only possible explanation for the finding had lung deposition, clinical response, plasma levels and side-effects been measured in the same subjects in the same study. Interestingly, a second deposition study [27], carried out using a different model of the Spacehaler compared with that used in the earlier deposition study [26] showed significantly higher lung deposition from Spacehaler than from pMDI.

Individual studies provide the most compelling evidence for the link between the deposition of inhaled asthma drugs and their clinical effects, and can be subdivided into two types of investigation: (i) studies where an improvement in drug delivery to the lungs with a particular inhaler device can be directly related to equivalent control of asthma with a lower dose of drug when that device is used, and (ii) studies where the efficacy of an inhaled drug is directly correlated with its lung deposition for two or more treatment regimens, and a series of case studies involving one or other of these situations will now be reviewed. Studies where lung deposition and pharmacodynamic effects have been assessed together in the same study provide the strongest evidence, although relatively few such studies have been conducted.

Double the lung deposition: halve the dose

Case study 1: pMDI vs pMDI plus spacer

Plastic spacer devices such as the Nebuhaler (AstraZeneca) are often used in conjunction with a pMDI, especially to deliver inhaled corticosteroids. Several gamma scintigraphy studies conducted in groups of healthy volunteers and patients using optimal inhaler technique have shown that when attached to a Nebuhaler, the percentage of the dose deposited in the lungs from a pMDI may be doubled or even trebled [28, 29]. There is some variability between the various scintigraphy studies in the mean amount of drug deposited in the lungs with the Nebuhaler, probably due mainly to different amounts of static charge, which attracts the aerosol droplets onto the spacer wall, so that they are not available for inhalation. A clinical trial conducted by Toogood et al. [30] assessed the effects of two dose levels (400 μg and 1600 μg per day over 2 week treatment periods) of budesonide from pMDI and pMDI plus Nebuhaler in 35 asthmatic patients, again using optimal inhalation technique, and concluded that a given level of antiasthmatic response could be achieved with approximately half the dose when the Nebuhaler was used. Hence the reduction in drug dose needed for a certain clinical response was correlated with improvement in delivery efficiency with the Nebuhaler (Table 2). In addition, there was significantly less oral candidiasis with the Nebuhaler [30], correlating with the reduction in oropharyngeal deposition which spacer devices such as Nebuhaler inevitably bring about [19].

Table 2.

Mean lung deposition (% ex-valve dose) from a pMDI and from a pMDI plus Nebuhaler spacer device. The mean increase in FEV₁ produced by 1600 μg daily from pMDI and pMDI plus Nebuhaler are in the same ratio as lung deposition values for the two products. Data from Newman et al. [28] and Toogood et al. [30].

Case study 2: pMDI vs Turbuhaler

Broadly similar data exist for the Turbuhaler dry powder inhaler (AstraZeneca) compared with a pMDI. Several studies involving both healthy volunteers and asthmatic patients have shown that a Turbuhaler approximately doubles the percentage of the dose deposited in the lungs compared with a pMDI [13, 29], when the two devices are compared in the same subjects and are used optimally (fast inhalation from Turbuhaler, and pMDI fired during a slow inhalation). For instance, the study by Thorsson et al. [29] showed mean lung depositions of 11.9% and 26.1% of the dose for pMDI and Turbuhaler, respectively, in a group of asthmatic patients. A placebo-controlled study in 50 patients with mild to moderate asthma conducted at seven centres compared salbutamol delivered in two dose levels from each device. The study showed that doses of 400 μg by pMDI and 200 μg by Turbuhaler were equipotent, and 100 μg by pMDI and 50 μg by Turbuhaler were equipotent [31]. The lower doses from each device gave a significantly smaller therapeutic response than the higher doses, indicating that they represented points further down the slope of the dose–response curve. The higher potency of salbutamol to produce bronchodilatation when given via the Turbuhaler was confirmed subsequently in a cumulative dose–response study [32]. It was concluded that the same bronchodilator effect could be achieved with half the dose of salbutamol given by Turbuhaler compared with pMDI, correlating with the improved efficiency of delivery to the lungs that the Turbuhaler produces.

Case study 3: pMDI vs Respimat

The Respimat (Boehringer) is a multiple dose liquid spray device (or ‘soft mist inhaler’) that may offer a viable alternative to pMDIs and dry powder inhalers in the maintenance therapy of asthma. The Respimat delivers very small liquid droplets which are better able to penetrate into the lungs than the spray from a pMDI. Comparative studies in a group of healthy volunteers with a radiolabelled formulation of the β-adrenoceptor agonist fenoterol showed that a mean of 14% of the dose was deposited in the lung for pMDI compared with 31% lung deposition for Respimat [33]. Likewise, with the corticosteroid flunisolide lung deposition in another group of healthy volunteers averaged 15% for pMDI compared with 39% for Respimat [34]. Inhalation from both devices in these studies involved firing the inhalers during slow deep inhalation, followed by 10 s breath holding. Comparative clinical trials have shown that with Respimat, only half the amount of drug is needed to give the same effect as by pMDI [35, 36]. In the study by Vincken et al. [36], the bronchodilator efficacy and safety of fenoterol was assessed in 41 asthmatic patients for Respimat (50 μg per puff), Respimat (100 μg per puff) and pMDI (100 μg per puff), with cumulative doses up to 12 puffs being given on each of 3 study days in randomised order. The data showed that the bronchodilator effects of fenoterol given by Respimat as 50 μg per puff were equivalent to those for pMDI given as 100 μg per puff, and there was a trend for 50 μg per puff by Respimat to give a lower increase in serum potassium and less finger tremor. Hence once again the improvement in lung deposition from a novel device permitted a lower drug dose to be used in order to maintain the therapeutic effect of an inhaled asthma drug.

Deposition vs response correlated directly

Case study 4: pMDI

The deposition from a pMDI of radiolabelled aerosol in the lungs under a variety of carefully controlled inhalation techniques (involving slow and fast inhalations, and short and long breath holding pauses) has been assessed in asthmatic patients, and has been compared with the bronchodilator response to a standard dose (500 μg) of the β-adrenoceptor agonist terbutaline sulphate inhaled from a pMDI under the same carefully controlled conditions. Optimal inhaler technique (slow deep inhalation plus 10 s breath holding) maximized lung deposition and clinical response, while other inhalation modes (faster inhalation and/or shorter breath hold) reduced lung deposition and bronchodilator response [37]. The optimal inhalation technique yielded points on the top of the dose–response curve, while other inhalation techniques gave points on the slope of the curve.

Case study 5: pMDI vs Turbuhaler

The lung deposition and clinical response of two dose levels of terbutaline sulphate (250 μg and 500 μg) were assessed in a 4-period cross-over study in 13 patients with moderate asthma following inhalation from pMDI and Turbuhaler [38]. The charcoal block method was used to quantify lung deposition, which averaged 8.1% and 8.3% of the nominal dose for pMDI, and 19.0% and 22.0% for Turbuhaler. The FEV₁ response after inhaling 250 μg terbutaline sulphate from Turbuhaler was significantly greater than that after inhaling the same dose from pMDI. Doses of 250 μg and 500 μg terbutaline sulphate by Turbuhaler, and 500 μg terbutaline sulphate by pMDI, gave responses at or close to the top of the dose–response curve, while 250 μg by pMDI gave a response on the slope of the dose–response curve (Table 3). It was concluded that the observed differences in drug deposition were reflected in the bronchodilator responses to inhaled terbutaline sulphate.

Table 3.

Correlation between lung dose of terbutaline sulphate and peak increase in forced expiratory volume in one second (FEV₁) in 11 asthmatic patients following inhalation from pMDI or Turbuhaler dry powder inhaler. Lung doses greater than about 40 emsp14;μg led to a maximal or near-maximal bronchodilator response. Data replotted from Borgström et al. [37].

Case study 6: sodium cromoglycate inhaled at different flow rates

Deposition of sodium cromoglycate (SCG) in the lungs was correlated with the protective effects of this drug against allergen challenge induced bronchoconstriction in an elegant study by Laube et al. [39]. Eight asthmatic patients (baseline FEV₁ 84% predicted) underwent inhaled allergen challenge following pretreatment with SCG that had been radiolabelled with the radionuclide ^99mTc, and which was inhaled at either slow (30 l min⁻¹) or fast (70 l min⁻¹) inhaled flow rates. Slow inhalation resulted in higher whole lung deposition of drug, a more peripheral drug distribution (lower value of inner lung zone deposition expressed relative to outer lung zone deposition), and a more homogenous distribution within the lungs (quantified via a lower value of skew in the aerosol distribution). There was a correspondingly smaller decrease in FEV₁ following allergen challenge when SCG was inhaled at the slow inhaled flow rate (Table 4), which was ascribed to the improved deposition of drug at its target site in the lungs.

Table 4.

Mean (s.d.) total and regional lung deposition of sodium cromoglycate at slow and fast inhalation rates, and subsequent mean (s.d.) fall in FEV₁ after allergen challenge. Data from Laube et al. [38].

Case study 7: pMDI vs Autohaler

The deposition of 100 μg salbutamol radiolabelled with ^99mTc from a pMDI was assessed in eight asthmatics using (a) the patients’ usual inhaler techniques (b) taught inhaler technique, and (c) an Autohaler device which fired the spray automatically during inhalation, thus ensuring ‘coordination’ between firing and inhaling [40]. Taught pMDI technique and Autohaler resulted in lung depositions averaging 20.8 and 22.8% of the dose, and caused about 30% bronchodilatation, while patients’ usual techniques resulted in lung deposition averaging only 7.2%, with only about 10% bronchodilatation. Several individuals fired the pMDI after completing their inhalation, thus depositing negligible drug in their lungs, and failing subsequently to achieve a significant bronchodilator response.

Discussion

The considerations outlined above suggest that whole lung deposition data are predictive of the clinical response to inhaled asthma drugs, and raise the possibility that lung deposition data could be used to replace clinical response data under appropriate circumstances as a means of demonstrating the effectiveness of new inhaled drug products to the regulatory authorities. There would be clear cut benefits to the pharmaceutical industry from using lung deposition data instead of clinical response data as pivotal data in regulatory submissions, primarily related to significant savings of both time and development costs.

These benefits are particularly marked in the cases of drugs such as inhaled corticosteroids which do not bring about a rapid and easy-to-measure clinical response. For instance, the use of lung deposition data instead of clinical efficacy data for generic inhaled corticosteroids products could be used to show bioequivalence between inhaled products with a much smaller population size in a two-way cross-over study on single study days, compared with a large-scale 4 week parallel group study used to assess clinical response directly [41, 42]. Lung deposition data could also be used instead of clinical efficacy data to document the effects of changes in manufacturing site or manufacturing process, as has been suggested elsewhere [43]. For new inhaled drug products, it is probably impossible to avoid the conduct of large scale phase III clinical efficacy studies, but a lung deposition study intended to show comparability between the new product and an established ‘gold standard’ product could be conducted instead of a phase II dose-ranging study, and we are aware of instances on both sides of the Atlantic where this has occurred.

The situations described above are ones in which either equivalence or comparability between test and reference products is to be expected. The use of lung deposition data as a surrogate for clinical response becomes harder to justify when two products with very different drug delivery characteristics are being compared. For instance, it has been shown that the lung deposition of a non-CFC formulation of beclomethasone dipropionate (Qvar, 3m Healthcare) is approximately 10 times that of a CFC formulation [44], while the ratio of drug doses required for satisfactory asthma control is approximately 2.0 [45]. The non-CFC formulation comprised very small aerosol droplets and gave a much more peripheral distribution pattern within the lungs than the CFC formulation, but it was not possible to quantify the relative amounts of drug delivered to the receptor sites from each product.

Gamma scintigraphy will generally be the method of choice in situations where lung deposition data can act as a surrogate for the clinical response to inhaled asthma drugs [6]. Currently, there is lack of standardization between methodologies at different centres performing scintigraphic studies, so that differences in the quantification of lung deposition are likely to result. Moves to standardize scintigraphic methodology, begun by the BALR Consensus Statement [6], and continued by the International Society for Aerosols in Medicine should solve this difficulty. The appropriate use of lung deposition data in regulatory packages should lead to significant time saving in the drug development process, and hence a faster development programme for inhaled asthma products.