Professional Documents
Culture Documents
Owen 2012
Owen 2012
REVIEW ARTICLE
Drug Safety R&D, Pfizer Global R&D, Sandwich, Kent, United Kingdom
Abstract
The preclinical safety studies required to support the development of inhaled drugs are generally the same as for
other routes of administration. Repeat-dose toxicology studies should be conducted by inhalation to ensure the
characterization of both the local (i.e., respiratory) and systemic toxicity, although some studies (e.g., reproductive)
can be performed by utilizing alternative routes, when it is paramount to maximize systemic exposure. Respiratory
tract changes in preclinical species can include irritancy of the larynx and nasal cavity, particularly in rodents. Such
changes are not necessarily predictive of a risk to humans because of the exquisite sensitivity of the rodent larynx
and the lack of exposure to the nasal cavity after oro-inhalation of drugs in the clinical setting. The design of poorly
soluble molecules to limit systemic exposure places greater emphasis on the elimination of drugs from the lungs by
macrophages. Consequently, an increase in macrophage numbers is often noted, and in the absence of any other
changes, this is generally considered to be a nonadverse, physiological response to an inhaled particulate. Other
For personal use only.
changes in the lung, which can include an inflammatory response and/or epithelial hyperplasia, resulting from
irritancy or particulate overload, are a safety concern and are not monitorable in humans. For such changes, safety
margins can be calculated in terms of the drug deposited per unit weight of lung. These factors should be taken into
account when designing preclinical studies or programs for inhaled drugs.
Keywords: Preclinical, inhalation, risk assessment, pharmaceutical, lung, pathology, toxicokinetics, contamination
1
2 K. Owen
Table 1. Current inhaled pharmaceutical agents. effects in the respiratory tract, in addition to defining the
Use Class of drug Examples systemic toxicity of the drug. In some cases, it may be
Asthma/COPD SABA Salbutamol necessary to supplement the inhaled dose with an oral or
LABA Salmeterol parenteral dose to attain systemic exposures high enough
Formoterol to induce toxicity (DeGeorge et al., 1997), as was the case
Indacaterol during the development of salmeterol (Owen et al., 2010).
LAMA Ipratropium When studying the carcinogenicity of an inhaled drug,
Tiotropium at least one of the two rodent studies should utilize the
ICS Beclomethasone inhaled route of administration (DeGeorge et al., 1997).
Fluticasone Although the oral route can be used as an alternative,
Mometasone it would not be suitable if proliferative or preneoplastic
Ciclesonide changes (e.g., hyperplasia) had been noted in the respi-
LABA/ICS Salmeterol/ ratory tract in chronic inhaled toxicity studies, there was
combination fluticasone no distribution to respiratory tract tissues, or if the meta-
Formoterol/ bolic profile after oral administration was significantly
Drug and Chemical Toxicology Downloaded from informahealthcare.com by University of Calgary on 06/17/12
hypertension Iloprost
Anesthetics Halogenated Isoflurane be designed to be of low solubility and oral bioavailabil-
hydrocarbons Halothane ity. Therefore, to maximize exposure, these studies are
COPD, chronic obstructive pulmonary disease; SABA, short- often performed using the i.v. (or subcutaneous) route
acting β-agonists; LABA, long-acting β-agonist; LAMA, long- of administration; conversely, no such issue would exist
acting muscarinic antagonist; ICS, inhaled corticosteroids; for inhaled drugs designed for the treatment of systemic
rDNA, recombinant DNA.
diseases.
respiratory infections are also well established for use by Method of administration
this route of administration (see Table 1). The basic concept of administration of drugs to labora-
Newer strategies have evolved that target the delivery tory animals by inhalation from a dynamic system is
of systemically active pharmaceuticals through the lungs, quite simple. A test aerosol is generated into a stream of
and this has led to inhaled insulin preparations and air, which is drawn through a chamber by an exhaust/
drugs in development to treat pain and migraine among pump system. The animals can either be exposed within
other conditions (Rubkin, 2010; Wadher et al., 2011). large chambers (i.e., whole body) or the drug can be
Drugs designed for the treatments of such conditions delivered, by significantly smaller chambers, direct to
are usually soluble and bioavailable, therefore allowing the breathing zone (Drew, 1990; Wong, 2007). The use
rapid systemic drug delivery through deep lung inhala- of whole body exposure systems in the pharmaceutical
tion. Once inhaled, the drug is quickly distributed into industry has diminished as a result of the large quanti-
the bloodstream, providing a fast onset of action that is ties of test material required and the potential for cross-
comparable to i.v. administration, but with greater ease, contamination and ingestion through grooming. Rodent
patient comfort, and convenience. and lagomorph studies usually employ a snout-only
The package of preclinical safety studies needed to exposure system, whereas nonrodent studies are gener-
support the administration of inhaled drugs to humans ally performed using an oro-nasal mask (Wong, 2007).
in clinical trials is generally the same as that for other Administration through an oro-pharyngeal tube can
routes of administration, with development plans usu- also be used for large animal species when the supply
ally following the recommendations outlined in the rel- of test article is very limited, a short exposure time is
evant International Conference on Harmonization (ICH) required, or when dose-limiting nasal irritancy has been
guidelines (e.g., ICH, 2009, 2011). Where possible, repeat- encountered.
dose toxicology studies should employ the inhaled route Using the dosing systems available, inhalable atmo-
of administration to mimic the intended clinical route of spheres can be successfully generated from metered
administration and to ascertain any potential for adverse dose inhalers, dry powder blends, and nebulized
new methods of particle engineering are therefore ing phase of a study using an appropriate analytical
being developed for inhaled biopharmaceuticals, such method, whereas total particulate concentrations are
as microcrystallization, spray drying, and supercriti- determined gravimetrically. These extra procedures
cal fluid technology (Shoyele and Cawthorne, 2006). ultimately make the studies more labor intensive and
In these circumstances, specially developed devices therefore more expensive.
are sometimes needed for clinical use; however, they The design principles for inhaled toxicology studies
are generally unsuitable for use in toxicology studies are the same as for any other route of administration
because of the requirement for a high concentration and should include an evaluation of both systemic and
and/or exposure period, leading to the need to adapt or local (i.e., respiratory tract) effects of the drug. Acute
construct entirely new devices for the toxicology work studies are generally not performed and regulatory
(Hardy, 2008). requirements for the assessment of extremely high
The deliberate generation of an aerosol and the wide doses can be obtained from dose-escalating or dose-
For personal use only.
range of doses, combined with the very low limits of quan- range–finding studies, which are performed to aid in
tification (often in the pg/mL range) required to detect the selection of the high doses for subsequent subacute
the drug in biological fluids at very low doses, means that and chronic repeat-dose studies. Studies should include
cross-contamination between dose groups (including both sexes and should be performed in one rodent and
controls) can occasionally occur and would be apparent one nonrodent species. Exposure should occur daily,
during toxicokinetic assessments. However, a variety of along with the examination of animals for clinical signs
procedures to limit transfer of drug between groups by of toxicity. Periodic examinations should include body
equipment and clothing, along with the use of individu- weight, food consumption, clinical pathology (e.g.,
ally ventilated cages, and the use of cabinets with good hematology, blood biochemistry, and urinalysis), oph-
air extraction, which are maintained at a slightly negative thalmoscopy, electrocardiography (e.g., nonrodent),
pressure, can be employed to avoid excessive contami- and toxicokinetics.
nation and potential invalidation of a study. The use of Information on systemic exposure of animals during
naïve controls, analysis of plasma for metabolites, and repeated-dose toxicity studies is essential for the inter-
measurement of control chamber concentrations can all pretation of the results (ICH, 1995; EMEA, 2005). The
be used to provide reassurance that the control animals principles involved for determining the toxicokinetics
have not received the test material in error. of a drug during inhaled studies are the same as for
Intratracheal instillation is not a method that should any other route of administration, except that in some
be used routinely to characterize the toxicity of an circumstances, extremely low doses (e.g., <10 μg/kg)
inhaled drug. Although it may be used to deposit a pre- may be utilized and the dose is estimated rather than
cise amount of compound directly into the lungs, the known. The first sample is often taken after the end of
method is nonphysiological, involving invasive delivery the exposure period to avoid the potential for sample
and anesthetized animals, whereas the distribution of contamination, whereas the systemic exposure is often
the test compound throughout the lungs may not be normalized to take into account the daily variability of
uniform and may not replicate the same pulmonary delivered dose and so to give a value more represen-
distribution after inhalation. Further, the dose and/or tative of the overall achieved doses over the course
dose rate are usually substantially greater than would be of the study. Determination of lung concentrations,
observed after inhalation and the upper respiratory tract and therefore an indication of the percent deposition,
is bypassed (Driscoll et al., 2000). This method, however, can be performed using homogenized tissues and
can be useful in providing information on the relative suitable “cold methods,” such has high-performance
toxicity between compounds or over a range of doses liquid chromatography/tandem mass spectrometry mass
(Wong, 2007; Pauluhn, 2008). spectrometry. Single-dose radiolabel studies can be
along with the bronchial lymph node, should also be and F = inhaled fraction (%).
examined. Of these parameters, airstream concentration, body
In some cases, it may be desirable to use broncho- weight, and exposure time will be known. Some labo-
alveolar lavage (BAL) analysis to complement histo- ratories have the capability to determine RMV during a
pathologic examination of the lung. This can be used study; however, it is usually an estimate based on a func-
to identify early indicators of biochemical changes in tion of body weight according to the formula developed
circumstances where it is desirable to investigate the by Bide et al. (2000) or by the Association of Inhalation
mechanism of the toxicity. The fact that these indicators Toxicologists (Alexander et al., 2008). The latter method
can be quantified also provides a means of developing is gaining popularity because it utilized RMV measure-
a dose-response curve for the inhalation toxicity of the ments from laboratory animal species only, whereas
drug of interest and may provide a “no effect” level of Bide et al. utilized 18 animal species ranging from mice
exposure for regulatory purposes. Obviously, time- to giraffes. For simplicity, the inhaled fraction is often
course studies in rodents would require additional assumed to be 100%.
animals, and if lung lavage is performed, the unlavaged Dry powder blends of drug and lactose (with or
lobe should be sectioned for histopathology at multiple without ternary agents) are currently the preferred for-
levels (Henderson, 2005). The total and differential cell mulation for the development of drugs to treat pulmo-
count in BAL fluid is a good measure of lung inflamma- nary diseases. For practical (e.g., reproducibility and
tion. The presence of neutrophils is a sensitive indicator overloading of equipment) and scientific (e.g., avoiding
of an inflammatory response, whereas an increase in excessive lung deposition, increased particle size result-
lymphocytes or eosinophils suggests an immune-based ing from aggregation, and obstruction at the proximal
response. There are also a number of early biochemical nose as a result of dust overload) reasons, the maximum
mediators of pulmonary inflammation, which include achievable doses are often limited so as not to exceed a
tumour necrosis factor-alpha and interleukin-1, which total particulate concentration of 2 mg/L (Whalan and
are cytokines released from the resident macrophages Redden, 1994). Under such circumstances, high dose
that promote the adherence of circulating inflamma- levels in the region of 40–50 and 20–25 mg/kg can be
tory cells to the endothelium, whereas an increased achieved when using a 1-hour exposure and a 40% (w/w)
protein content can suggest an increased permeability dry power/lactose blend in rats and dogs, respectively.
of the alveolar/capillary barrier. Other markers of lung Achieving these doses is dependent on a variety of factors
toxicity include lactate dehydrogenase, a cytoplasmic in addition to the blend strength and exposure period,
enzyme whose presence extracellularly indicates cell such as the efficiency of the apparatus, chamber extract
death, beta-glucuronidase activity, which is a measure flow rate, generator speed, and surface area of the packed
is debateable, although in the case of nebulized studies tracheobronchial tract by the process of sedimentation,
with poorly soluble drugs, prolonged exposures can be with particles of 1 μm and less reaching the alveoli and
necessary to provide sufficient lung margins to allow a being deposited by diffusion (Inchiosa, 2006).
robust evaluation of local tolerance. Species differences in regional particle deposition
The achievable doses for solutions will primarily do exist, with obligate nose-breathing rodents show-
depend on the solubility of the drug substance, but also ing higher nasal and tracheobronchial deposition than
on exposure period, efficacy of the apparatus, nebulizer pulmonary deposition, compared to that observed in
airflow input and output, and the chamber aerosol extrac- humans. There is experimental evidence to suggest that
tion rate. A strategy can be employed during the develop- inhaled particles of a MMAD between 1 and 4 μm will
ment of a dry powder formulation to perform the initial deposit within all regions of the rat respiratory tract, but
toxicology and clinical studies with a nebulized solution, within this size range, nasopharyngeal and tracheobron-
because this can result in a rapid transition from selec- chial deposition increase as particle size increases, with
For personal use only.
tion of a drug candidate to the clinic. This strategy avoids pulmonary deposition remaining relatively constant.
the prolonged dry powder formulation work required Once particles are larger than 4 μm, deposition is pri-
before toxicology can commence and the drug require- marily in the nasopharyngeal region (Raabe et al., 1988;
ments and doses (e.g., <1 mg/kg for a 1-hour exposure) Schlesinger, 1985), with pulmonary deposition dropping
are relatively low as a result of the relatively insoluble off rapidly so that exposure at 6 μm is small and at 10 μm
nature of the material, thus minimizing the time needed negligible. For dogs and monkeys, the deposition profiles
for chemical synthesis. However, this strategy will not are much more like those observed for humans. Thus,
necessarily characterize any potential for local irritancy for inhalation toxicology studies, an aerosol bracket-
with the final dry powder formulation and so can lead ing a MMAD of 1–4 (for acutes) or 1–3 μm (for repeated
to overall longer times to significant clinical trials (e.g., administration) is recommended, with a geometric stan-
phase IIb) while additional studies are performed to dard deviation (GSD) in the range of 1.5–3.0 (SOT, 1992;
address this gap. In some cases, there is also a need to OECD, 2009b). Some organizations use 7 μm as a cutoff
change salt form because of difficulties in finding stable for exposure to the lungs, and indeed if the MMAD was
crystalline forms suitable for formulation as a dry powder. >7 μm, then the study would be considered to be invalid
Doses achieved for metered dose inhalers are generally with little or no pulmonary exposure occurring and thus
influenced by the same generic system limitations as for one of the major aims of an inhaled study (i.e., detection
other formulations, but are further affected by how many of local adverse effects) not being achieved.
actuations are possible simultaneously (e.g., six), the rate Clearance of particles from the respiratory tract occurs
at which the canisters can be discharged (e.g., 10/min- by a combination of mechanisms. Particulate deposited
ute), and the shot strength (e.g., 50–300 μg); these factors in the nasopharynx region may be removed by coughing
usually result in doses of less than 1 mg/kg in rodents. or sneezing, or is swallowed. Clearance from the tra-
cheobronchial region is primarily along the mucociliary
Deposition and clearance escalator system to a point where the particles can be
A substantial proportion of an orally inhaled dose expectorated or swallowed. Although material that is
is deposited in the oropharynx and swallowed, thus swallowed is removed from the respiratory tract, it is liable
becoming subject to oral absorption. How much drug to be absorbed into the systemic circulation, depending
is absorbed through the gastrointestinal tract would on its bioavailability/solubility. Another important clear-
depend on the proportion swallowed and the oral bio- ance mechanism for those materials reaching the tra-
availability of the drug. For salbumatol, it has been cheobronchial tree or alveoli is by the phagocytic activity
estimated that 60–80% is swallowed, with only 10–20% of macrophages. These particle-laden macrophages then
reaching the lungs after administration from a meter can move up the mucociliarly escalator or be removed
whereas, despite variation in specific isoenzymes, the tory response and loss of cilia will eventually stimulate
total levels in rats are similar between the two tissues squamous metaplasia of the mixed ciliated and noncili-
(Turton and Hooson, 1998). In addition, a number of ated (i.e., transitional) epithelium. The epithelium can
reportedly lung-specific CYPs, such as CYP2S, 2F, and eventually become hyperplastic and -keratotic, being
4B1, exist (Somers et al., 2007). Generally, the highest covered by a thick stratum corneum (Renne and Gideon,
levels of CYP isoenzymes appear to be located in the 2006). Studies have demonstrated partial or complete
olfactory epithelium of the Bowman’s gland in the nasal regression of laryngeal squamous metaplasia and hyper-
cavity, in Clara cells, and, to a lesser extent, type II cells plasia after recovery periods (Osimitz et al., 2007), with
of the lung (Dahl, 1995; Dormons and Van Bree, 1995). complete recovery taking between 6 and 13 weeks after a
Given the metabolic capability of the lung, production 13-week exposure period (Renne et al., 2007). There has
of reactive metabolites in lung cells could contribute to been no report of such changes progressing to more seri-
lung-specific toxicities (Dahl and Gerde, 1994). ous cellular changes (Osimitz et al., 2007).
For personal use only.
observed with inhaled pharmaceuticals, because drugs as they clear insoluble material. This phenomenon does
or doses that are severely irritant would result in either a not occur with soluble drugs or vehicles (e.g., lactose)
reduction in dose as the compound progresses through because they do not need to be cleared by macrophages.
the drug development cycle or discontinuation of the
project before commencement of longer term studies as Risk assessment
a result of small safety margins. A thorough review of the As for other routes of administration, safety margins for
many pulmonary changes that can be seen in laboratory behavioral changes or target organ toxicity after the inha-
animals after inhalation of irritant material is contained lation of a drug should be expressed in terms of multiples
in the publications by Gopinath et al. (1987) and Greaves of the systemic exposure at the maximum recommended
(2007). therapeutic dose in humans. Safety margins for target
organ toxicities (e.g., hepatotoxicity and nongenotoxic
Macrophage accumulation and lung overload carcinogenicity) are generally based on systemic expo-
For personal use only.
One consequence of designing drugs that have a very sures of the drug in terms of plasma area under the curve,
low solubility, to avoid systemic absorption, is significant whereas for acute effects (e.g., tachycardia and convul-
amounts of undissolved drug accumulating in the alveoli. sions), the margins are often based on maximum plasma
One of the normal clearance mechanisms is for these par- concentrations. In the absence of toxicokinetic data,
ticles to be engulfed by macrophages and then cleared to exposure margins should be expressed in terms of body
the exterior on the mucociliary escalator or by migration surface area (mg/m2).
to the pulmonary interstitium and, subsequently, the In terms of upper respiratory tract changes (e.g.,
local lymph nodes (Snipes, 1995; Miller, 2000). Evidence squamous metaplasia), the rodent larynx is extremely
of such normal clearance of dust is frequently visible sensitive to insult from inhaled particles and the
in lung sections from control animals as aggregates of changes are not usually predictive of irritancy in
foamy macrophages, typically located at the bronchiolar- humans (Gopinath et al., 1987; Osimitz et al., 2007).
alveolar junction, with a lack of any other associated Lesions resulting from irritancy in the nasal cavity are
changes (Lee et al., 1986). Often, an increase in the inci- largely considered irrelevant to a drug administered to
dence of this finding is noted in treated groups; however, humans by oro-inhalation, because little or no exposure
such changes are considered not adverse and are known would occur in the clinical setting (Dudley et al., 1989;
to be fully reversible (Warheit et al., 1997; Driscoll et al., Owen et al., 2010). Additionally, it should be noted that
1996). Occasionally, such changes are accompanied by irritant effects on both these regions of the upper respi-
an increase in lung weight (Elder et al., 2005). If the rate ratory tract would be easily monitorable in humans
of deposition exceeds the maximum rate of clearance receiving such a drug by inhalation. Also, whereas such
and/or dissolution, the insoluble drug accumulates in lesions are often noted in rodents, their presence in
the macrophages and interstitium until it causes impair- dogs or monkeys is relatively rare (Renne et al., 2007).
ment of macrophage clearance function and therefore a If required, safety margins can be calculated for nasal
further increase in the rate of accumulation, leading to changes using standard nasal surface area values and
lung overload (Morrow, 1992; Snipes, 1995). The induc- particle sizes, which would be determined in the toxi-
tion of these interrelated events has been postulated to cology study.
be related to the volume of material phagocytized by the Adverse pulmonary changes, which can include
alveolar macrophage pool (Morrow, 1992). The thresh- cellular infiltration, epithelial degeneration, necrosis,
old for inert particles causing load overload has been hyperplasia, and fibrosis, are considered to be a safety
reported to be approximately 1 mg/g of lung (Morrow, concern because they are adverse and nonmonitorable
1986). A progressive neutrophilic inflammatory reaction in humans and so require appropriate safety margins.
then occurs in response to the increasing particle load. In calculating safety margins for lung changes, the drug
that lung safety margins should be a minimum of 10-fold although restraint and thermal stress has been sug-
for rodents and 5- to 6-fold for nonrodents. Though the gested to play a role (Lee et al., 1993; Dodd et al., 1999).
preclinical deposition values can be refined using “real” Regardless of the cause, the snout-only conditions are
data generated in the toxicology studies (see study design not relevant to the human situation.
section above), it is unclear whether data from humans To summarise, though general principles for deter-
obtained by gamma scintigraphy, positron emission mining drug safety are the same for all routes of admin-
tomography, single-photon emission computed tomog- istration, there are some peculiarities resulting from
raphy (Newman et al., 2003), or direct (Nave et al., 2010) administration of drugs to animals by inhalation (e.g.,
or indirect pharmacokinetic methods (Thorsson et al., dose estimation, prolonged exposure periods, lack of
1994) would influence the view of the FDA because sensitive biomarkers of pulmonary toxicity, and use of
their deposition assumption is based on philosophi- lung deposition factors). These should be considered
cal safety concerns, rather than experimentally derived in the design of preclinical studies and the interpreta-
For personal use only.
deposition of inhaled monodisperse coarse and fine aerosol Accessed 5th June 2011.
particles in small laboratory animals. Ann Occup Hyg 32:53–63. Wadher, K., Kalsait, R., Umekar, M. (2011). Pulmonary insulin delivery:
Renne, R. A., Gideon, K. M. (2006). Types and patterns of response in challenges and current status. J Pharmaceut Sci Res 3:1052–1059.
the larynx following inhalation. Toxicol Pathol 34:281–285. Warheit, D. B., Hansen, J. F., Yuen, I. S., Kelly, D. P., Snajdr, S. I., Hartsky,
Renne, R. A., Gideon, K. M., Harbo, S. J., Staska, L. M., Grumbein, S. M. A. (1997). Inhalation of high concentrations of low toxicity dusts
L. (2007). Upper respiratory tract lesions in inhalation toxicology. in rats results in impaired pulmonary clearance mechanisms and
Toxicol Pathol 35:163–169. persistent inflammation. Toxicol Appl Pharmacol 145:10–22.
Rubkin, B. K. (2010). Air and soul: the science and application of Webber, C. (2005). Inhaled-dose ADME studies. Dev Life Sci 6:10–13.
aerosol therapy. Respir Care 55:911–921. Whalan, J. E., Redden, J. C. (1994). Interim policy for particle size
Schlesinger, R. B. (1985). Comparative deposition of inhaled aerosols and limit concentration issues in inhalation toxicity sudies.
in experimental animals and humans: a review. J Toxicol Environ U.S. EPA. Available at: cfpub.epa.gov/ncea/cfm/recordisplay.
Health 15:197–214. cfm?deid = 186068. Accessed September 2011. 2nd Sept 2011.
Shoyele, S. A., Cawthorne, S. (2006). Particle engineering tech Wilson, A. (1987). Aerosol dynamics and delivery systems. In: Jenne,
niques for inhaled biopharmaceuticals. Adv Drug Deliv Rev J. W., Murphy, S. (Eds.), Drug therapy for asthma: research and
58:1009–1029. clinical practice (pp. 389–412). New York: Marcel Dekker.
For personal use only.
Snipes, M. B. (1989). Long-term retention and clearance of particles Witschi, H., Last, J. A. (2001). Toxic responses of the respiratory system.
inhaled by mammalian species. Crit Rev Toxicol 20:175–211. In: Klassen, C. (Ed.), Casarett and Doull’s toxicology, 6th ed. (pp.
Snipes, M. B. (1995). Pulmonary retention of particles and fibers. 515–534). New York: McGraw-Hill Medical Publishing Division.
In: Concepts in inhalation toxicology, 2nd ed. (pp. 193–227). Wolff, R. K. (1998). Safety of inhaled proteins for therapeutic use. J
Washington, DC: Taylor and Francis. Aerosol Med 11:197–219.
Society of Toxicology (SOT). (1992). Technical committee of the Wong, B. A. (2007). Inhalation exposure systems: design, methods,
Inhalation Speciality Section. Society of Toxicology (SOT) and operation. Toxicol Pathol 35:3–14.
recommendations for the conduct of acute inhalation limit tests. World Heath Organization. (2008). Factsheet 310: top 10 causes
Fundam Appl Toxicol 18:321–327. of death. Available at: www.who.int/mediacentre/factsheets/
Somers, G. I., Lindsay, N., Lowdon, B. A., Jones, A. E., Freathy, C., Ho, fs310_2008.pdf. AccessedApril 2011. 14th April 2011.
S., et al. (2007). A comparison of the expression and metabolizing Young, S. L., Silbajoris, R. (1986). Dexamethasone increases adult rat
activities of phase I and II enzymes in freshly isolated human lung lung surfactant lipids. J Appl Physiol 60:1665–1672.