Pharmacology and Therapeutics of Bronchodilators

A. A History of the Development of β-Adrenergic Receptor Agonists: from Nonselective Adrenergic Receptor Agonists to β₂-Adrenergic Receptor-Selective Drugs

In traditional Chinese medicine, the botanical ma huang (the plant Ephedra equisetina), from which the active material, an alkaloid identified as ephedrine, is extracted, was used for more than 2000 years for the short-term treatment of respiratory symptoms. Beginning at the turn of the last century, the nonselective α-AR and β-AR agonist epinephrine was introduced into clinical practice and administered by the subcutaneous route for the treatment of acute asthma (Bullowa and Kaplan, 1903). Although highly efficacious, epinephrine was far from the ideal drug, because in addition to the desired therapeutic effect of bronchodilation, it caused hypertension and tachycardia, now known to be due to its effect on α₁- and β₁-ARs in the cardiovascular system. Moreover, because of metabolic instability, it had a short duration of action and had to be given parenterally (Waldeck, 2002). In the 1940s, the nonselective β-AR agonist isoproterenol was introduced for the treatment of airway disease and became the standard-of-care bronchodilator, although its use was complicated by adverse effects that were due to activation of the β₁-AR in the heart (Waldeck, 2002), which elicits tachycardia and predisposes patients to cardiac dysrhythmias. A further disadvantage of isoproterenol is its short duration of action, which is due to its ready transportation into cells by the uptake process for catecholamines (Gryglewski and Vane, 1970) where, except in the gut, it is converted by catechol-O-methyltransferase (COMT) to 3-O-methyl-isoprenaline (Blackwell et al., 1974). This limits the duration of effect, and the compound has a poor oral availability (Waldeck, 2002). Metaproterenol, a noncatechol resorcinol derivative of isoproterenol, was subsequently developed in the early 1960s. It was an effective bronchodilator when inhaled but also did not discriminate between β₁- and β₂-ARs and thus produced cardiac side effects. The modern era of selective β₂-AR agonists did not begin until the discovery of albuterol (called salbutamol in Europe) by Sir David Jack and colleagues working at Allen and Hanburys in the UK (now part of GlaxoSmithKline) (Jack, 1991).

All the clinically important β₂-AR agonists consist of a benzene ring with a chain of two carbon atoms and either an amine head group or a substituted amine head group. If a hydroxyl (OH) group is present at positions 3 or 4 on the benzene ring, the structure is a catechol nucleus and hence the agent is termed a catecholamine. If these OH groups are substituted or repositioned, the drug is generally less potent than the synthetic catecholamine isoproterenol, which has both strong β₁- and β₂-AR properties and is the more powerful bronchodilator (McFadden, 1981). However, this potential disadvantage may be outweighed by the relative resistance of substituted catecholamines to metabolic degradation by COMT. The noncatecholamine β₂-AR agonists such as fenoterol, albuterol, and terbutaline differ in their substitutions in the amine group and benzene ring. These structural modifications, conferring resistance to metabolism by COMT, result in a longer half- life and also reduce their potency for β₁-ARs, making them relatively more selective for β₂-ARs (Fig. 3).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Milestone development of β2-AR agonists (A), antimuscarinics (B), and xanthines (C).

A major limitation of the β₂-AR agonists in use during the 1960s and 1970s was their short duration of action, typically 4 to 6 h. Therefore, the next advance in the development of β₂-AR agonists was the development of the long-acting drugs salmeterol and formoterol, the duration of action of which is approximately 12 h, which made their use for maintenance treatment (e.g., for reducing nighttime symptoms) more appealing (Lötvall, 2002). A pure R-isomer of albuterol, levalbuterol (Berger, 2003), and the R,R-enantiomer of formoterol, arformoterol (Cazzola et al., 2010b), have been developed. It is claimed that they have a better safety profile than the racemic mixture because they do not have the S-enantiomer, which, at least for (S)-albuterol, is now known to have unwanted effects in the lung (Page and Morley, 1999). At present, several once-a-day ultra-long-acting β₂-AR agonists are in different stages of clinical development (Cazzola and Matera, 2009; Cazzola et al., 2011) and will be discussed in more detail in section II.D.

B. Short-Acting β₂-Adrenergic Receptor Agonists

Short-acting β₂-AR agonists (SABAs) can be divided into two broad groups according to duration of action after inhalation of conventional doses: 1) the catecholamines isoprenaline and rimiterol, which have a very short action of 1 to 2 h; and 2) those conventionally described as short-acting, such as fenoterol, albuterol, and terbutaline, which are active for 3 to 6 h, although the action of fenoterol may be slightly longer (Beardshaw et al., 1974). In any case, the duration of action of all β₂-AR agonists increases with dose such that 1600 μg of inhaled albuterol, for example, has a considerably longer action than 200 μg (Corris et al., 1983).

C. Long-Acting β₂-Adrenergic Receptor Agonists

Long-acting β₂-AR agonists (LABAs) such as salmeterol and formoterol provide 12-h bronchodilation (Lötvall, 2002). The duration of action of β₂-AR agonists in the human bronchus is in the following order: salmeterol ≫ formoterol ≥ albuterol ≥ terbutaline > fenoterol.

Although formoterol and salmeterol are both potent and effective β₂-AR agonists, their different chemical structures confer markedly different pharmacological characteristics. It has been suggested that salmeterol specifically binds to the β₂-AR via the albuterol “head group,” whereas a secondary exosite has been proposed in which the lipid tail binds to confer the long duration of action (Green et al., 1996). The exosite is an auxiliary binding site, a domain of highly hydrophobic amino acids within the fourth domain of the β₂-AR (Johnson, 1998). When the lipid tail is in association with the exosite, the salmeterol is prevented from dissociating from the β₂-AR, but the head can freely engage and disengage the active site by the Charniére (hinge) principle, the fulcrum being the oxygen atom in the side chain. The position of this oxygen atom is critical for the long duration of action (Johnson, 2001). The salmeterol molecule is >10,000 times more lipophilic than albuterol. It partitions rapidly (<1 min) into the cell membrane and then diffuses laterally to approach the active site of the β₂-AR through the membrane. The process is relatively slow (>30 min) and accounts for the slow onset of action of salmeterol compared with albuterol (Johnson, 2001).

On the contrary, formoterol is not thought to be able to access the exosite and an alternative hypothesis for its long duration of action has been proposed whereby the lipophilic, basic nature of this drug allows formoterol to partition effectively into the lipid bilayers of ASM after inhalation. This partitioning then permits an effective concentration of agonist to be present over time in the form of a depot within the ASM, from where formoterol progressively leaches out to interact with the active site of the β₂-AR, providing a prolonged duration of action. This is known as the diffusion microkinetic theory (Anderson et al., 1994). The size of the depot is determined by the concentration or dose of formoterol applied. In airway preparations, the onset of action of formoterol is somewhat delayed compared with albuterol, and the duration of relaxant activity, although longer, is concentration-dependent (Johnson, 2001). In any case, formoterol is somewhat less lipophilic than salmeterol and is believed to diffuse more rapidly through the lung tissues, reaching the site of action faster than salmeterol (Anderson, 1993).

D. Ultra-Long-Acting β₂-Adrenergic Receptor Agonists

A variety of β₂-AR agonists with longer half-lives are currently in development, with the hope of achieving once-daily dosing (Cazzola et al., 2005b, 2011; Matera and Cazzola, 2007; Cazzola and Matera, 2008, 2009). These agents include indacaterol (which received European regulatory approval in November 2009 and has already been launched in several countries worldwide), olodaterol, vilanterol, carmoterol, 5-(2-{[6-(2,2-difluoro-2-phenylethoxy)hexyl]amino}-1(R)-hydroxyethyl)-8-hydroxyquinolin-2(1H)-one (LAS100977), N-((4′-hydroxybiphenyl-3-yl)methyl)-2-(3-(2-((-2-hydroxy-2-(4-hydroxy-3-((methylsulfonyl)amino)phenyl)ethyl)amino)-2-methylpropyl)phenyl)acetamide (PF-610355), and AZD-3199, the structure of which has not yet been disclosed but has been suggested by Norman (2009).

E. Intravenous β₂-Adrenergic Receptor Agonists

An interesting new option is the development of a β₂-AR agonist to be administered intravenously. Bedoradrine (MN-221) is a novel, highly selective β₂-adrenoceptor agonist under development for the treatment of acute exacerbation of asthma and COPD. Bedoradrine was calculated to be 832- and 126-fold more selective for β₂-ARs than for β₁- and β₃-ARs, respectively, which indicates that it is very highly selective for β₂-ARs in this assay system (Inoue et al., 2009). However, it was initially considered for the treatment of preterm labor because it is 1590-fold more specific for the uterus than for the trachea (Inoue et al., 2009).

The PK and pharmacodynamics (PD) of bedoradrine were investigated using data from a single intravenous dose study in patients with stable moderate to severe COPD (Sadler et al., 2010). Patients receiving doses of 600 and 1200 μg showed responses superior to those of patients receiving 300 μg. At 1200 μg, the mean peak FEV₁ increase was approximately 55% of maximal lending support to this dose.

Although both albuterol and bedoradrine independently induced an increase in heart rate, adverse effects on heart rate were not observed upon combination in dogs. Other cardiovascular parameters (QTc and monophasic action potential) were not adversely affected in the albuterol + bedoradrine combination. These findings are consistent with some other data implicating bedoradrine as a partial agonist at β₁-ARs (Johnson et al., 2010).

A study that evaluated the safety and tolerability of bedoradrine at doses of 150 to 900 μg via intravenous infusion in patients with mild to moderate stable asthma documented that it was safe and well tolerated, with dose-dependent improvements in FEV₁ (Matsuda et al., 2010). A preliminary small trial showed that bedoradrine added to standard therapy for severe acute asthma exacerbations was safe and provided additional clinical benefit (Nowak et al., 2010).

In a small group of patients with COPD, bedoradrine, at doses of 300, 600, or 1200 μg i.v., seemed to improve lung function at all dose levels and reached statistical significance at both 600 and 1200 μg compared with placebo (Pearle et al., 2010). FEV₁ (L) increased compared with baseline by an average of 21.5% (P = 0.0025) for the 1200-μg dose, 16.2% (P = 0.02) for the 600-μg dose, and 9.2% (P = NS) for the 300-μg dose compared with a decrease of 4.0% for placebo. Bedoradrine was generally well tolerated by all patients.

F. Side Effects of β₂-Adrenergic Receptor Agonists

Because of the widespread distribution of β₂-ARs, a number of undesired responses result when β₂-AR agonists are absorbed into the systemic circulation. Therefore, side effects of β₂-AR agonists are greatest when the drugs are administered orally or parenterally (Svedmyr and Löfdahl, 1996), but there is some risk of cardiac actions even with inhaled formulations (Cazzola et al., 1998c, 2005a).

Increased heart rate and palpitations are less common with the selective β₂-AR agonists than with nonselective β₁- and β₂-AR agonists, although even stimulation of β₂-ARs can result in vasodilation and reflex tachycardia. Furthermore, some of the β-ARs in the atria and ventricles are β₂ in type, and thus direct simulation of the heart results from the use of even selective β₂-AR agonists (Cazzola et al., 1998c, 2005a). Consequently, they should be used with caution in patients with hyperthyroidism or cardiovascular disease (arrhythmias, hypertension, QT-interval prolongation) (Cazzola et al., 1998c, 2005a).

Administration of β₂-AR agonists to patients with airways obstruction often results in a transient decrease in partial pressure of oxygen in arterial blood (PaO₂) despite concomitant bronchodilation. This transient decrease in PaO₂ has been attributed to the pulmonary vasodilator action of these agents (Knudson and Constantine, 1967) as a result of the activation of β₂-ARs that are present in pulmonary blood vessels (Conner and Reid, 1984), increasing blood flow to poorly ventilated lung regions and thus increasing ventilation-perfusion (VA/Q) inequality, a shunt-like effect, at least in patients with asthma (Wagner et al., 1978). Hypoxemia induced by β₂-AR agonists is usually mild (Cazzola et al., 2006; Santus et al., 2007). Intriguingly, there is evidence that, in more vulnerable patients with severe COPD undergoing exacerbations, nebulized albuterol does not aggravate pulmonary gas exchange. In contrast, there is a small deterioration in the gas exchange response to albuterol in the same patients whose baseline lung function has improved as a result of increased perfusion of low-V/Q-ratio regions, while these patients are in convalescence (Polverino et al., 2007).

β₂-AR agonists should also be used with caution in patients with diabetes because of the risk of ketoacidosis. β₂-AR stimulation in the liver induces glycogenolysis and therefore raises blood sugar levels (Philipson, 2002). Hypokalemia is also a risk because of stimulation in skeletal muscle of the Na⁺,K⁺-ATPase-driven pump coupled to β₂-ARs. This increases the ability of the Na⁺,K⁺ pump to push Na⁺ out of the cell and facilitate intracellular accumulation of K⁺, thereby lowering plasma levels. An increase in blood flow to the skeletal muscles through β₂-AR-mediated vasodilation could also contribute to the increase in skeletal muscle K⁺ levels (Tesfamariam et al., 1998). Numerous studies have shown a dose-related fall in serum K⁺ level with increasing doses of β₂-AR agonist (Scheinin et al., 1987), although there is some evidence that this effect shows tolerance and that serum K⁺ levels after regular treatment are not reduced (Canepa-Anson et al., 1987). Such hypokalemia may precipitate arrhythmias; hence, the use of β₂-AR agonists is linked to an increased incidence of tachyarrhythmias (Du Plooy et al., 1994).

A further side effect of selective β₂-AR agonists is a fine tremor of skeletal muscle, particularly of the hands, which can be inhibited in experimental animals by the selective β₂-AR antagonist (±)-1-[2,3-(dihydro-7-methyl-1H-inden-4-yl)oxy]-3-[(1-methylethyl)amino]-2-butanol (ICI 118551) (Tesfamariam et al., 1998). It is greatest after oral or parenteral routes of administration (Ahrens, 1990). An early explanation of the tremor was that β₂-AR stimulation shortens the active state of skeletal muscle, which leads to incomplete fusion and reduced tension of tetanic contractions (Waldeck, 1976). Normally, tetanic contractions involved in maintaining steady muscle tone are fused. This property, together with β₂-AR-mediated enhancement of skeletal muscle spindle discharge, could explain the development of muscle tremor often experienced by patients using β₂-AR agonists. More recently, the tremor has been correlated closely with the hypokalemia, and it was suggested that the tremor was a direct result of the raised intracellular K⁺ levels in skeletal muscle (Tesfamariam et al., 1998). However, tolerance to the tremorogenic effects of β₂-AR agonists occurs with their long-term use (Ahrens, 1990).

β₂-AR agonists can induce the mobilization of triglycerides resulting in elevated blood levels of fatty acids and glycerol (Smith and Kendall, 1984), although it is the β₁-AR that is responsible for mediating this effect. Albuterol, terbutaline, and fenoterol can induce mild appetite suppression, headache, nausea, and sleep disturbances (Miller and Rice, 1980; Pratt, 1982). This is consistent with their ability to cross the blood-brain barrier, leading to CNS levels approximately 5% of those seen in plasma (Caccia and Fong 1984). Use of indacaterol can also elicit cough in some users, although when cough does occur, it is mild in severity, transient in nature and duration, and declines with the duration of treatment (Cazzola et al., 2010d).

A. Brief History of Development

Inhaled mAChR antagonists have been used as treatments for respiratory diseases for centuries. The smoking of plant alkaloids was recommended as a therapy for asthma in the literature of Ayurvedic medicine as early as the 17th century (Gandevia, 1975; Jackson, 2010). Atropa belladonna and Datura stramonium are rich in anticholinergic alkaloids such as atropine and stramonium (Miraldi et al., 2001). These plants can be dried and smoked as a mechanism to deliver these agents to the airways. The practice was introduced to Great Britain in the early 19th century by travelers from India who had observed its beneficial effects. It was soon adopted widely, and for the next century and a half, stramonium and belladonna in the form of burning powders, cigarettes, and cigars were used throughout Europe and North America in the inhalation treatment of asthma. Surveys and clinical studies conducted during the 1940s and 1950s continued occasionally to emphasize the therapeutic value of cigarettes containing either stramonium or atropine.

Unfortunately, atropine, which is a tertiary ammonium compound, is well absorbed into the systemic circulation and penetrates the blood-brain barrier. As a result, it has multiple systemic side effects that limit its clinical usefulness. For this reason, it slowly lost popularity. However, a renewed interest in anticholinergic drugs as therapy for respiratory diseases has been sparked by the development of safe yet effective quaternary anticholinergic compounds (Fig. 3). Chemical modifications of the atropine molecule, in particular by making its nitrogen atom pentavalent, have yielded a number of synthetic congeners that are very poorly absorbed from mucosae and cross the blood-brain barrier with difficulty, but in other respects they have pharmacological properties similar to those of atropine (Cazzola et al., 1998a; Gross, 2006). The quaternary structure of these anticholinergics prevents a substantial absorption from mucosal surfaces and penetration of the blood-brain barrier. When given by inhalation, these agents are as effective as atropine at improving lung function but longer acting and much less prone to side effects (Gross, 2006).

B. Short-Acting Muscarinic Acetylcholine Receptor Antagonists

As well as atropine methonitrate, ipratropium bromide and oxitropium bromide are effective short-acting quaternary anticholinergic drugs that have been used in the treatment of respiratory diseases. Their duration of action is approximately 6 to 8 h, but compared with SABAs, they have a slower onset of action, although probably a longer duration of action (Cazzola et al., 1998c; Gross, 2006).

C. Long-Acting Muscarinic Acetylcholine Receptor Antagonists

D. Novel Long-Acting Muscarinic Acetylcholine Receptor Antagonists

The growing evidence of the importance of tiotropium bromide in the maintenance treatment of COPD has spurred further research to identify new long-acting mAChR antagonists that could be as successful as tiotropium bromide (Cazzola and Matera, 2009). Several novel inhaled LAMAs are currently being developed, including glycopyrronium bromide (NVA-237), aclidinium bromide, umeclidinium bromide (GSK573719), TD-4208, 3-((((3-fluorophenyl)((3,4,5-trifluorophenyl)methyl)amino)carbonyl)oxy)-1-(2-oxo-2-(2-thienyl)ethyl)-1-azoniabicyclo(2.2.2)octane (CHF 5407), QAT370 (structure not disclosed), BEA-2180BR (structure not disclosed), trospium, dexpirronium, and PF-4522971 (structure not disclosed).

We still do not know whether these new mAChR antagonists are really as long-lasting as tiotropium bromide or whether they have a longer duration of action simply as a consequence of increasing the dose to overcome their shorter duration of action, possibly through a mechanism implying the generation of a drug depot in the lungs. Recently, however, it has been documented that with the same total daily dose of a new mAChR antagonist, a twice-daily regimen provides higher bronchodilation at trough than the once-daily regimen (Wu et al., 2011). For the maximum bronchodilation, there is a small difference between the two regimens, with the once-daily regimen being slightly better than the twice-daily regimen. For the overall bronchodilation response, as quantified by the FEV₁ response 24-h area under the curve, the difference between the two regimens becomes even smaller.

E. Side Effects of Muscarinic Acetylcholine Receptor Antagonists

All of the currently approved inhaled mAChR antagonists have a very wide therapeutic margin and are very well tolerated, in part because they are very poorly absorbed after inhalation. However, if any of these agents makes inadvertent contact with the eye, they can cause papillary dilation and blurred vision (Kizer et al., 1999), a true risk in patients with glaucoma (Mulpeter et al., 1992). Angle-closure glaucoma precipitated by ipratropium bromide after both nebulizer and pressurized aerosol treatment (Malani et al., 1982) and by tiotropium bromide (Oksuz et al., 2007) has been reported. Angle closure glaucoma could be attributed to a possible angle crowding as a result of pupillary dilation caused by the inhibitory effect of mAChR antagonists on the parasympathetic nervous system. This effect on the eye is particularly important in the case of tiotropium bromide, because its prolonged duration of action may precipitate acute glaucoma.

In older men, who may have prostatic hyperplasia, mAChR antagonists should be used with caution because they can cause urinary retention (Pras et al., 1991). A nested case-control study of patients with COPD aged 66 years or older documented that men who just initiated a regimen of inhaled mAChR antagonists were at increased risk for acute urinary retention compared with nonusers (Stephenson et al., 2011). In men with evidence of benign prostatic hyperplasia, the risk was further increased. Men using both short- and long-acting inhaled mAChR antagonists had a significantly higher risk of acute urinary retention compared with monotherapy users or nonusers. Patients with moderate-to-severe renal impairment (creatinine clearance of ≤50 ml/min) under treatment with tiotropium bromide should be closely monitored, because tiotropium bromide is predominantly excreted by the kidneys through active secretion (Keam and Keating, 2004). A bad taste and dryness of the mouth has been reported by 20 to 30% of patients, with a frequency in tiotropium bromide that may be slightly greater than that observed with ipratropium bromide (Gross, 2004).

Paradoxical bronchoconstriction to ipratropium bromide has been reported in humans (Mann et al., 1984). It is possible that this results from blockade of prejunctional M₂ mAChRs on airway cholinergic nerves, which normally inhibit ACh release; when the drug is given by nebulizer, however, it is largely explained by the hypotonicity of the nebulizer solution and by antibacterial additives, such as benzalkonium chloride. Paradoxical bronchoconstriction, which may also occur with other mAChR antagonists, warrants withdrawal of the drug from the patient.

Concerns have been raised about possible associations of mAChR antagonists with cardiovascular morbidity and mortality (Lee et al., 2008; Singh et al., 2008b). However, the results of the UPLIFT trial and a robust and extensive analysis of more than 19,000 patients participating in placebo-controlled clinical trials with tiotropium bromide indicate that there is not a real increased risk for death or cardiovascular morbidity during treatment with this inhaled anticholinergic agent in patients with COPD (Tashkin et al., 2008; Celli et al., 2010). Nonetheless, the results of a recent systematic review and meta-analysis indicate that tiotropium bromide mist inhaler is associated with a statistically significant increased risk of mortality, possibly because this new formulation and devise delivers the drug more effectively to the lungs and leads to greater absorption into the systemic circulation (Singh et al., 2011b).

A. History of Development

One of the earliest reports of the efficacy of methylxanthines in asthma was published in the Edinburgh Medical Journal in 1859, where Henry Hyde Salter, himself an asthmatic, described his experience that “one of the commonest and best reputed remedies of asthma… is strong coffee” (Salter, 1859). The first analysis of a xanthine derivative extracted from tea leaves was accomplished by Kossel (1888), who was able to extract not only caffeine but also another xanthine derivative, a dimethylxanthine. The name theophylline was then applied to a compound that has two methyl groups (1–3 dimethylxanthine) (Mazza, 1982; Schultze-Werninghaus and Meier-Sydow, 1982) (Fig. 3). Until 1930, xanthine derivatives were used in clinical practice only because of their diuretic and cardiotonic properties, even though the original studies, which clearly demonstrated the antispasmodic action of theophylline on smooth muscle of bronchi and its superiority in the relief of attacks of asthma, were carried out independently by Macht and Ting (1921) in the United States and Hirsch (1922) in Germany. Years later, a combination of theophylline and ethylenediamine (aminophylline) was used intravenously as an effective bronchodilator in acute asthma (Herrmann et al., 1937). Since then several other xanthines have been synthesized and are used clinically in various parts of the world.

B. Theophylline

Theophylline is a methylxanthine similar in structure to the common dietary xanthines caffeine and theobromine. It directly relaxes human ASM in vitro, but it is a relatively weak bronchodilator with an effective concentration giving 50% response of 1.5 × 10⁻⁴ M in vitro, which equates to a plasma concentration of 67 mg/l, assuming 60% protein binding (Guillot et al., 1984). However, the therapeutic range of plasma concentrations has been established to be 10 to 20 mg/l because of unacceptable side effects above plasma concentrations of 20 mg/l (Barnes, 2003), although there is also evidence that maintaining levels of theophylline as low as 5 mg/l provides good anti-inflammatory activity (Sullivan et al., 1994). The dose of theophylline required to give therapeutic concentrations varies among patients, largely because of differences in clearance of the drug. However, theophylline is an effective bronchoprotective agent at therapeutic levels (Magnussen et al., 1987) and is also able to inhibit airway hyper-responsiveness (Ward et al., 1993; Page et al., 1998). Theophylline has been used for the treatment of asthma for many years. Prolonged treatment with theophylline was associated with a significant change in the sensitivity and slope of the methacholine dose-response curve that was qualitatively similar to that observed after regular treatment with corticosteroids, albeit to a lesser extent (Page et al., 1998).

Theophylline is rapidly and completely absorbed, but there are large interindividual variations in clearance because of differences in hepatic metabolism. It is eliminated by the hepatic cytochrome P450 system (85–90%) and urinary excretion (10–15%). Theophylline is predominantly metabolized by the CYP1A2 enzyme, whereas at higher plasma concentrations, CYP2E1 is also involved (Zhang and Kaminsky, 1995). Diet, concomitant cardiac or liver disease, tobacco use, and several medications affecting the cytochrome P450 system can modify the half-life of theophylline (Boswell-Smith et al., 2006). Theophylline is approximately 40% protein bound in plasma, the remaining portion circulating throughout the body. Disease states that may affect protein binding, such as end-stage liver disease, renal dysfunction, and hypoalbuminemia, may result in a higher concentration of unbound drug.

It has been observed that FEV₁, FVC, and peak expiratory flow rate changed only slightly (approximately 13%) over the range of doses that induced steady-state serum theophylline concentrations of 5 to 10, 10 to 15, and 15 to 20 mg/l, respectively (Chrystyn et al., 1988). Theophylline can also act as a respiratory stimulant. Moreover, theophylline is able to improve diaphragmatic contractility and has anti-inflammatory properties. There is, in fact, good evidence for inhibitory effects of theophylline on airway inflammation in patients with asthma (Sullivan et al., 1994) and COPD, and these effects are seen at plasma concentrations below 10 mg/l (Kobayashi et al., 2004).

Theophylline has been widely used in the treatment of patients with asthma and COPD. Many studies have examined its efficacy in these patients, but results are conflicting and difficult to interpret because of differences in patient populations, duration of therapy, and measurements of outcome. Apparently, the withdrawal of theophylline can worsen lung function, clinical status, exercise performance, and ratings of dyspnea (Kirsten et al., 1993). Studies comparing theophylline with β₂-AR agonists found similar improvements in spirometry and in various measures of patients' functional state and well-being (Cazzola et al., 1999b). However, very few studies have compared theophylline with inhaled mAChR antagonists.

Despite its extensive use in the treatment of respiratory disease, the precise molecular action(s) of theophylline have not been fully elucidated. Its efficacy in the treatment of patients with COPD or asthma has traditionally been attributed to nonselective phosphodiesterase (PDE) inhibition (Nicholson and Shahid, 1994), resulting in an increase in cAMP by inhibition of PDE3 and PDE4 and an increase in cGMP by inhibition of PDE5. However, it has been demonstrated that theophylline administered at therapeutic plasma levels (10 mg/l) exerts only a very modest nonselective inhibition of PDEs, which might not be clinically relevant (Barnes, 2003); at the lower doses now recommended, which have been shown to have significant anti-inflammatory activity, there is little evidence that theophylline inhibits any of the known PDE enzyme families (Boswell-Smith et al., 2006). Thus a number of other potential mechanisms of action of theophylline have been proposed, including adenosine receptor antagonism (Yasui et al., 2000) and increasing histone deacetylase 2 (HDAC2) activity (Ito et al., 2002). Intriguingly, theophylline restores HDAC2 activity to normal in macrophages obtained from patients with COPD and reverses corticosteroid resistance (Cosio et al., 2004). This can explain why mice exposed to cigarette smoke develop corticosteroid-resistant inflammation that is reversed by low doses of oral theophylline (Fox et al., 2007; To et al., 2010). The molecular mechanism of action of theophylline in restoring HDAC2 seems to be via selective inhibition of phosphoinositide 3-kinase δ, which is activated by oxidative stress in patients with COPD and in smokers with asthma (Marwick et al., 2010; To et al., 2010). Although the side-effect profile of xanthines may relate to adenosine receptor antagonism (particularly A₁ receptor), it is unlikely that the therapeutic benefit of xanthines occurs via this mechanism because enprophylline has been shown to be as effective as theophylline in the treatment of patients with asthma, yet this drug lacks adenosine receptor antagonism (Persson and Pauwels, 1991).

C. Aminophylline

Aminophylline is the ethylenediamine salt of theophylline with higher solubility at a neutral pH. In vivo intravenous aminophylline has an acute bronchodilator effect in patients with asthma that is most likely to be due to a relaxant effect on ASM (Mitenko and Ogilvie, 1973). However, it should not be given to a patient taking oral theophylline until the serum theophylline level has been measured. In any case, there is no evidence to support the use of intravenous aminophylline in acute severe asthma in adults (Hart, 2000). In critically ill children with severe acute asthma, aminophylline can confer early benefits on airway function and oxygenation, sustained for 24 h for oxygenation but not for airway function, and reduces the risk of endotracheal intubation (Yung and South, 1998). However, side effects such as nausea and vomiting are common, and the benefit/risk ratio is unknown. Aminophylline is also widely used for the treatment of acute exacerbations of COPD despite a lack of evidence to support this practice (Brochard et al., 1995). It remains possible that aminophylline confers a small but clinically significant benefit when added to standard therapy. However, Duffy et al. (2005) were unable to find evidence for any clinically important additional effect of aminophylline treatment when used with high-dose nebulized bronchodilators and oral corticosteroids in patients with nonacidotic COPD exacerbations. Aminophylline also increases diaphragmatic contractility and reverses diaphragm fatigue (Aubier et al., 1981).

E. Novel Xanthine Drugs

The positive clinical effects of theophylline in airway disease, combined with its advantageous oral bioavailability, has spurred the development of other xanthines, such as bamiphylline, acebrophylline, and doxofylline, for the treatment of respiratory disease, with the anticipation that such drugs would have greater efficacy than theophylline but with an improved side-effect profile.

The structural novelty in doxofylline arises because of the unique methylene 1,3-dioxolone substitution at the N7 position of the xanthine ring. It has greatly decreased affinity toward adenosine A₁ and A₂ receptors compared with theophylline, which may contribute to the better safety profile, because in a number of studies, doxofylline has been shown to have better efficacy and fewer side effects than theophylline (Page, 2010). Moreover, unlike theophylline, doxofylline does not interfere with calcium influx into cells or antagonize the action of calcium-channel blockers (Page, 2010). As a consequence, the effective therapeutic dose of doxofylline has fewer cardiostimulant effects than theophylline, such that doxofylline does not significantly increase the cardiac frequency or have arrhythmogenic effects (Page, 2010). Recent data also suggest that doxophylline does not directly inhibit any of the known HDAC enzymes and has much reduced ability to antagonize adenosine receptors or inhibit any known PDE enzyme (van Mastbergen et al., 2011). However, despite its superior safety efficacy, the potential of doxofylline has not been fully exploited.

A patented once-daily doxofylline tablet has been developed. This new sustained release formulation contains 650 mg of doxofylline and was shown to be bioequivalent to the standard recommended dose of 400 mg b.i.d. (Page, 2010). This novel formulation also shows lower peak plasma concentrations, thus widening the therapeutic window further; it is anticipated that this once-daily administration of 650 mg of doxofylline to patients with asthma or COPD will provide increased compliance and dosage convenience (Page, 2010).

F. Side Effects of Xanthines

Theophylline has a narrow therapeutic index; serum levels slightly outside of the target range may lead to serious toxicity or lack of efficacy. Long-term theophylline overmedication generally results from its sometimes unpredictable elimination kinetics [particularly in those at the extremes of age (i.e., very young and very old)], erratic use by patients, and administration to patients with underlying hepatic or cardiac disease.

Serum theophylline levels can be used to predict which patients are at risk for toxicity. However, life-threatening events are not predictable from very high serum theophylline levels, making it difficult to identify patients in whom aggressive interventional therapy is warranted before toxicity arises (Aitken and Martin, 1987). That theophylline concentrations are not predictive of toxic effects in patients with long-term overmedication has been recognized for many years (Shannon, 1999). This unusual feature of long-term overmedication can be explained by alterations in theophylline PK. For example, the volume of distribution of theophylline is 0.4 to 0.7 l/kg. Long-term overexposure to the drug would increase tissue concentrations, resulting in a greater body burden of drug relative to serum concentration. The lack of association between theophylline serum concentrations and adverse clinical effects in patients with long-term overmedication complicates the identification of high-risk patients who might benefit from early hemodialysis or hemoperfusion. Chronologic age has been increasingly shown to have greater predictive value in identifying patients at high risk, with those older than 60 to 65 years most likely to develop major toxic effects.

The most commonly reported adverse effects associated with theophylline include anorexia, nausea, headache, and sleep disturbance. Theophylline was also found to worsen symptoms of gastroesophageal reflux disease and to cause cardiac arrhythmias. Given that most of these adverse effects can be avoided at lower plasma levels, therapeutic drug level monitoring plays a pivotal role. However, levels within the therapeutic range should not preclude dosage decreases when a theophylline-related adverse event is suspected (Paul et al., 2010).

Several significant drug interactions can affect both the initial dose and subsequent dosing of theophylline. Because theophylline is metabolized by cytochrome P450 1A2 and to a lesser extent 3A3 and 2E1, medications that are inducers or inhibitors of these enzymes can affect the metabolism and elimination of theophylline. Concurrent administration of phenytoin and phenobarbitone increases activity of P450, resulting in increased metabolic breakdown, so that higher doses may be required. In addition, medications that are substrates of the same P450 enzymes, such as quinolone antibiotics (ciprofloxacin, but not ofloxacin), allopurinol, cimetidine (but not ranitidine), serotonin uptake inhibitors (fluvoxamine), and the 5-lipoxygenase inhibitor zileuton, which interferes with CYP1A2 function, will compete with theophylline for metabolism and potentially elevate theophylline levels (Aronson, 2006).

Older adults have a higher risk of toxicity when taking theophylline because of concomitant diseases, especially cardiovascular disease, reduced drug clearance, and polypharmacy, with consequent potential drug-drug interactions. It has been suggested that theophylline should be considered as potentially inappropriate for the older patients unless its plasma levels can be closely monitored, because it has a narrow therapeutic index and because plasma levels of theophylline are sensitive to reduced clearance, underlying diseases, and drug interactions (De Smet et al., 2007).

Some of the side effects of theophylline (central stimulation, gastric secretion, diuresis, and arrhythmias) may be due to adenosine receptor antagonism (predominantly A₁ receptor antagonism), but the most common side effects of theophylline are nausea, gastrointestinal symptoms, and headaches, which may be due to inhibition of certain PDEs (e.g., PDE4 in the vomiting center) (Howell et al., 1990). It is therefore of interest that doxofylline, which lacks significant A₁ receptor antagonism and PDE inhibitory activity, has a much improved side-effect profile compared with theophylline (Page, 2010; van Mastbergen et al., 2012). When theophylline is used as a bronchodilator at doses that give plasma concentrations of 10 to 20 mg/l, side effects due to PDE inhibition and adenosine antagonism are relatively common and often lead to discontinuation of therapy. However, a considerable body of evidence now suggests that theophylline works as a bronchodilator and anti-inflammatory agent at 5 to 10 mg/l concentration and that the use of low doses of theophylline that give plasma concentrations of 5 to 10 mg/l largely avoids side effects, thus makes it unnecessary to monitor plasma concentrations (unless checking for compliance) (Sullivan et al., 1994; Barnes, 2005).

A. Selective Phosphodiesterase Inhibitors

Theophylline, long considered a nonselective PDE inhibitor, is known to be effective in the treatment of respiratory diseases; as described in section IV.B, the wider use of this drug is limited by its narrow therapeutic window (Boswell-Smith et al., 2006; Diamant and Spina, 2011). There has therefore been a longstanding interest in finding more selective PDE inhibitors, and this has become more plausible with the identification over the past decade of 11 different isoenzymes (PDE1–11) within the PDE superfamily that catalyze the breakdown of the signaling molecules cAMP and/or cGMP. Each isoenzyme has different tissue subdivision and different properties, enabling targeted therapy with potentially fewer (systemic) side effects (Boswell-Smith et al., 2006; Diamant and Spina, 2011).

In particular, in contrast to nonselective PDE inhibition, there is considerable interest in selectively targeting the PDE4 isoenzyme because it is the predominant isoenzyme in the majority of inflammatory cells, including neutrophils, which are implicated in the pathogenesis of COPD. PDE4 is also present in ASM, but selective PDE4 inhibitors have not shown acute bronchodilator activity in a variety of clinical trials (Boswell-Smith et al., 2006; Calverley et al., 2009). However, the selective PDE4 inhibitors are often considered members of a new class of potential bronchodilators, which is misleading, and they should be classified as novel anti-inflammatory drugs. Nonetheless, we will consider the evidence that PDE4 selective inhibitors can influence ASM function.

Theophylline is a weak and nonselective inhibitor of PDEs with less than 10% inhibition of PDE4, in comparison, for example, with roflumilast, the first PDE4 inhibitor to be approved for clinical use, which achieves 30 to 60% inhibition of PDE4 subtypes at clinically effective doses (Boswell-Smith et al., 2006). However, although PDE4 is one of the enzymes found in human ASM and a number of PDE4 inhibitors have been shown to improve lung function in patients with COPD after long-term administration (for review, see Boswell-Smith et al., 2006), no PDE4 inhibitor to date has shown any direct bronchodilator effect in the clinic. It is therefore thought that the impact of PDE inhibitors on lung function after long-term dosing clinically is secondary to the well described anti-inflammatory effects of these agents (for review, see Boswell-Smith et al., 2006). Certainly in vitro studies using isolated human bronchial tissues demonstrate that various PDE4-inhibitors are poor functional antagonists of spasmogen-induced bronchoconstriction and produce only modest relaxation of ASM directly (Diamant and Spina, 2011).

A number of selective PDE4 inhibitors have been investigated clinically for their effects on lung function. Cilomilast failed to induce an acute bronchodilator effect (Grootendorst et al., 2003) and when administered for 6 weeks in a dose-finding study in patients with moderate to severe COPD, although it significantly increased baseline FEV₁ by 11% by the end of the study (Compton et al., 2001). Several clinical trials with roflumilast have documented that this PDE4 inhibitor caused a small improvement in pre- and postbronchodilator FEV₁ (48 and 55 ml, respectively) in patients with more severe COPD, especially in those with a chronic bronchitis phenotype, those with recent exacerbations, and those requiring frequent rescue inhaler use (Calverley et al., 2009). These small changes in FEV₁ are not thought to be because of direct bronchodilation and occurred only after regular treatment. It is possible, as suggested by Celli (2006), that orally active PDE4 inhibitors such as cilomilast and roflumilast could reach the distal airways because of their systemic distribution, where decreasing inflammation around the small airways could result in more important changes in resting lung volumes than in absolute FEV₁. Another possible explanation for the improvements in lung function could be partly attributed to a modulation of sensory nerve function by PDE4 inhibitors. A consequence of reduced sensory input to the central nervous system would lead to a decrease in parasympathetic outflow to the airways and a reduction in contraction of ASM and submucosal gland secretion, both leading to improvements in airflow (Spina et al., 1995). In any case, it has been demonstrated that PDE4 inhibitors can relax inherent tone in isolated human bronchial muscle (Schmidt et al., 2000) and that the PDE4D variant 5 is the key physiological regulator of β₂-AR-induced cAMP turnover within human ASM (Billington et al., 2008).

Several other oral PDE4 inhibitors, such as oglemilast (GRC-3886), MEM1414 (structure not disclosed), and tetomilast are under development (Matera et al., 2011b), although to date, the therapeutic usefulness of oral PDE4 inhibitors is limited by their side effects, particularly effects on the gastrointestinal system (Boswell-Smith et al., 2006; Diamant and Spina, 2011).

As an alternative to oral administration, a number of groups are now investigating the topical application of PDE4 inhibitors as a possible way to improve their efficacy in the treatment of inflammatory airway diseases while minimizing side effects. 6-((3-((Dimethylamino)carbonyl)phenyl)sulfonyl)-8-methyl-4-((3-methyloxyphenyl)amino)-3-quinolinecarboxamide (GSK256066), SCH900182 (structure not disclosed) (Chapman et al., 2010), and 1-[[5-(1(S)-aminoethly)-2-[8-methoxy-2-(trifluormethyl)-5-quinolinyl]-4-oxazolyl]carbonyl]-4(R)-[(cyclopropylcarbonyl)amino]-l-proline, ethyl ester xinafoate salt are examples of inhaled PDE4 inhibitors under development. GSK256066 has shown modest therapeutic potential in patients with asthma (Singh et al., 2010), but to date there are no studies in the literature with this drug in patients with COPD. However, failures in clinical trials have been reported for three other inhaled PDE4 inhibitors: tofimilast (Danto et al., 2007), N-(3,5-dichloropyrid-4-yl)-(1-(4-fluorobenzyl)-5-hydroxy-indole-3-yl)glyoxylic acid amide (AWD 12-281) (Gutke et al., 2005), and UK500001 (structure not disclosed) (Vestbo et al., 2009).

As discussed above, it is clear that PDE4 inhibitors have minimal direct effects of clinical relevance on ASM. In contrast, it has long been appreciated that PDE3 is the prevalent isoenzyme in the ASM that leads to smooth muscle relaxation. Indeed, a number of PDE3 inhibitors have been evaluated clinically for their effects on bronchodilation and have been demonstrated to have some activity (Leeman et al., 1987; Fujimura et al., 1997; Myou et al., 2003). However, PDE3 is also an important isoenzyme in cardiac and vascular tissue, and so PDE3 inhibitors may have detrimental effects on the cardiovascular system (Torphy, 1998). Nonetheless, it is clear that a safe mixed PDE3/4 inhibitor would provide both bronchodilator and anti-inflammatory activity in one molecule, something that is very attractive. Indeed, attempts have previously been made to clinically evaluate inhaled mixed PDE3/4 inhibitors such as zardaverine (Brunnée et al., 1992), but at a dose that acutely improved lung function in patients with asthma, the swallowed portion of the inhaled medicine led to unwanted gastrointestinal side effects. However, a nonemetic mixed PDE3/4 inhibitor, 9,10-dimethoxy-2-(2,4,6-trimethylphenylimino)-3-(N-carbamoyl-2-aminoethyl)-3,4,6,7-tetrahydro-2H-pyrimidino(6,1-a)isoquinolin-4-one (RPL554), has been demonstrated (Boswell-Smith et al., 2006) that is both a bronchodilator and anti-inflammatory. This drug has already successfully undergone a number of phase 2 clinical trials in patients with either COPD or asthma (http://www.veronapharma.com). There are also some recent pilot data showing that PDE5 inhibitors such as sildenafil can have bronchodilator activities (Sousa et al., 2011).

B. K⁺ Channel Openers

In ASM cells, K⁺ channels, such as such as large-conductance, voltage-dependent Ca²⁺-activated K⁺ channels (K_Ca) or the ATP-dependent K⁺ potassium channels (K_ATP), play an important role in modulating contractile activity. Activation of these channels will cause cell hyperpolarization that should oppose Ca²⁺ entry through voltage-dependent Ca²⁺ channels, leading to smooth muscle relaxation. Consequently, K⁺ channel modulators may be of value in the treatment of chronic airway disorders. However, K_ATP openers, although effective in relaxing human airways in vitro, are not effective in treating asthma or COPD because they are more potent as vasodilators, which limits the dose that can be administered safely (Nardi et al., 2008).

Emerging attention can instead be seen in the patent literature for K_Ca channel openers. K_Ca channels regulate smooth muscle responses to both contractile agonists that elevate intracellular calcium and relaxant agonists that elevate cAMP (Jaggar et al., 1998). Freshly isolated human ASM cells express the large conductance K_Ca1.1 channel. In addition to K_Ca1.1, two other types of K_Ca channel occur in human cells: the low conductance SK_Ca family (K_Ca2.1–2.3) and the intermediate conductance channel K_Ca3.1. Preclinical data support this interest, although clinical proof of concept has not yet been obtained because of the lack of potent and selective K_Ca channel openers. In vascular smooth muscle cells, the large conductance channel (K_Ca1.1), leads to membrane hyperpolarization and the inhibition of voltage-gated L-type Ca²⁺ channels. This, in turn, reduces the amount of Ca²⁺ influx and thereby decreases smooth muscle cell tone (Perez-Zoghbi et al., 2009). However, the observations that voltage-gated Ca²⁺-channel inhibitors have no effect on agonist-induced Ca²⁺ signaling in ASM and are ineffective bronchodilators and that the deletion of K_Ca1.1 channels in transgenic mice reduced airway contractility all suggest that the contribution of membrane potential to regulation of [Ca²⁺]_i in ASM cells is minimal (Perez-Zoghbi et al., 2009). Consequently, the role of K_Ca1.1 in ASM cell contraction is still controversial. In any case, it remains to be determined whether an oral K_Ca channel activator could selectively interact with ASM without causing adverse effects. There is documentation that K_Ca3.1 is expressed by both normal and asthmatic human ASM (Shepherd et al., 2007), and it has been suggested that K_Ca3.1 mediates important biological effects in the ASM of subjects with asthma and its blockade might at least partially prevent ASM remodelling (Bradding and Wulff, 2009). The only K⁺ channel modulating agent under clinical development in an airway indication is senicapoc, an orally active K_Ca3.1 blocker that inhibits the late airway response and the development of bronchial hyper-responsiveness after allergen challenge in a sheep model of asthma (Bradding and Wulff, 2009).

C. Vasoactive Intestinal Peptide Analogs

VIP, one of the major peptide transmitters in the central and peripheral nervous systems, is abundantly present in the normal human lung, and VIP-immunoreactive nerves are found in the smooth muscle layer and glands of the airway and within the walls of pulmonary and bronchial vessels (Onoue et al., 2007). The effects of VIP are mediated through interaction with two common G protein-coupled receptors named VPAC1 and VPAC2 (Laburthe and Couvineau, 2002). The bronchodilatory effect of VIP in human bronchi is almost 100 times more potent than adrenergic dilation induced by isoproterenol (Wu et al., 2011). Unfortunately, clinical application of VIP has been limited for a number of reasons, including its short plasma half-life and problems with administration, because intravenous VIP will lead to profound bronchodilation but also vasodilation; after inhalation to reduce systemic exposure, VIP is subject to degeneration by proteases present in the lung lining fluid (Wu et al., 2011). As a consequence several peptidase-resistant VIP analogs, including Ac-His¹[Glu⁸,OCH₃-Tyr¹⁰,Lys¹²,Nle¹⁷,Ala¹⁹,Asp²⁵,Leu²⁶,Lys^27,28]-VIP (Ro 25-1392), Ac-His¹[Glu⁸,Lys¹²,Nle¹⁷,Ala¹⁹,Asp²⁵,Leu²⁶,Lys^27,28, Gly^29,30,Thr³¹]-NH₂ VIP (Ro 25-1553), and [Arg^15,20,21, Leu¹⁷]-VIP (IK312532), have been developed. Ro 25-1553 is 24 to 89 times more potent than VIP as a relaxant of the guinea pig trachea, and pulmonary responses evoked by histamine, leukotriene D₄, platelet-activating factor, and ACh are inhibited dose dependently by intratracheally instilled Ro 25-1553 (O'Donnell et al., 1994). However, in patients with asthma, it elicits a short-lasting bronchodilatory effect (Wu et al., 2011). IK312532 exhibits longer-lasting bronchodilation and suppression of the antigen-evoked infiltration of granulocytes in the bronchiolar mucosa compared with VIP, possibly because of higher stability against enzymatic digestion (Ohmori et al., 2004).

D. Rho Kinase Inhibitors

There is now substantial evidence that Rho kinase (RhoK) is involved in bronchoconstriction (Fernandes et al., 2007). RhoK can modulate smooth muscle contraction by multiple mechanisms. Agonists can activate RhoK via Ca²⁺-independent pathways involving G protein-coupled membrane receptors, subsequent activation of Rho family guanine nucleotide exchange factors and monomeric G protein RhoA, and eventual RhoK activation (Somlyo and Somlyo, 2003). RhoA, which is a member of the Rho (Ras-homologous) subfamily of the Ras superfamily, has been identified as the main upstream activator of RhoK. However, several studies, including some in airways, have shown that the RhoK pathway may also be activated by Ca²⁺-dependent mechanisms (Ito et al., 2001; Hirota et al., 2007). The contractile response to cholinergic stimulation of airways has been shown to depend partly on agonist-induced RhoK activation, which contributes to the early phase of the contractile response. RhoK modulates airway responsiveness via myosin light-chain phosphatase inhibition, without modifying the Ca²⁺ signal. RhoK is activated via a Ca²⁺-dependent pathway, but independently of Ca²⁺-calmodulin-dependent protein kinase II and Ca²⁺-activated Cl⁻ currents (Mbikou et al., 2011). Hence, the implication of RhoK signaling pathways in ASM is not fully determined, although it seems to be important.

Considering that Rho/RhoK signaling is thought to be involved in various processes that contribute to chronic airway diseases, the use of RhoK inhibitors in asthma and COPD therapy clearly holds promise. Inhibition of RhoK reduces contractile responses induced by spasmogens. Indeed, the RhoK inhibitor N-(4-pyridyl)-4-(1-aminoethyl)cyclohexanecarboxamide (Y-27632) has been shown to relax human isolated bronchial preparations (Gosens et al., 2004). Several analogs of Y-27632 exist that have similarly high inhibitory constants for RhoK and similar smooth muscle relaxant properties. Of those, 4-(1-aminoethyl)-N-(1H-pyrrolo(2,3-b)pyridin-4-yl)cyclohexanecarboxamide dihydrochloride (Y-30141) and Y-30694 (structure not disclosed) (Schaafsma et al., 2008) may be particularly worth mentioning, because their selectivity profiles with respect to PKC and myosin light chain kinase are only slightly dissimilar from that of Y-27632 (Gosens et al., 2006). Fasudil (HA-1077) has affinity for RhoK similar to that of Y-27632. A metabolite of fasudil, hydroxyfasudil (HA-1100) also relaxes smooth muscle with a potency similar to the parent compound. It is, however, far more potent (K_i for RhoK = 1.6 nM), and its selectivity profile (with respect to PKA, PKC, and myosin light chain kinase) is even better than that of Y-27632 (Schaafsma et al., 2008). Two aminofurazan-based inhibitors, N-(3-{[2-(4-amino-1,2,5-oxadiazol-3-yl)-1-ethyl-1H-imidazo[4,5-c]pyridin-6-yl]oxy}phenyl)-4-{[2-(4-morpholinyl)ethyl]-oxy}benzamide (GSK269962A) and 4-(7-{[(3S)-3-amino-1-pyrrolidinyl]carbonyl}-1-ethyl-1H-imidazo[4,5-c]pyridin-2-yl)-1,2,5-oxadiazol-3-amine (SB-7720770-B), have been identified as a novel class of potent inhibitors of ROCK activity. IC₅₀-values of GSK269962A and SB-7720770-B toward Rho-associated protein kinase 1 (ROCK-1) are 1.6 and 5.6 nM, respectively, and toward ROCK-2, 6 and 4 nM, respectively (Doe et al., 2007). Compared with Y-27632 and fasudil, both compounds were far more potent in inhibiting ROCK-1. In addition to ROCK-1 and -2 inhibition, SB-7720770-B potently inhibits several other kinases (e.g., mitogen- and stress-activated protein kinase 1 and ribosomal S6 kinase 1) as well, whereas GSK269962A is at least >30-fold more selective for ROCK compared with the protein kinases tested.

E. Brain Natriuretic Peptide and Analogs

It is well known that the guanylyl cyclase/cGMP second messenger system has a parallel role to the adenylyl cyclase/cAMP system in ASM, regulating its contractile and proliferative functions. Therefore, it is not surprising that agents activating this signaling pathway bronchodilate in vivo and relax ASM in vitro (Hamad et al., 1997). Particulate guanylyl cyclases act as plasma membrane receptors for natriuretic and related peptides. Several membrane forms of the enzyme have been identified up to now. Some of them serve as receptors for the natriuretic peptides, a family of peptides that includes atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and C-type natriuretic peptide, three peptides known to play important roles in renal and cardiovascular physiology. There is evidence that ANP relaxes human ASM in vitro (Angus et al., 1994b). Moreover, in humans, exogenous ANP reverses airway hyper-responsiveness when given intravenously or by inhalation in high doses, although it seems to be less effective when administered by inhalation (Angus et al., 1993, 1994a), and has also been shown to modify bronchial responsiveness to inhaled histamine (Angus et al., 1995) and nebulized water (Mcalpine et al., 1992). We have documented that BNP, which is able to induce a time- and concentration-dependent increase of cGMP levels in human ASM cells (Hamad et al., 2003), induces a weak but intriguing relaxant effect on human ASM (Matera et al., 2009) through the activation of the natriuretic peptide receptor-A (Matera et al., 2011a). The incubation of human ASM with BNP can elicit a significant reduction of cholinergic and histaminergic responses (Matera et al., 2011a). We speculate that the direct activation of the natriuretic peptide receptor-A might influence bronchial tone, inducing bronchodilation, and certainly it has been documented that human recombinant BNP (nesiritide) is a potent bronchodilator in patients with asthma (Akerman et al., 2006).

F. Nitrix Oxide Donors

Exogenous NO has the ability to exert bronchodilatory effects in patients with bronchial asthma (Högman et al., 1993), and NO has been used in the treatment of preterm children to improve lung capacity (Barrington and Finer, 2007). The augmented availability of NO in the lungs may represent a plausible approach for the treatment of asthma and COPD. Nitrates, NO, and NO donors relax ASM in vitro and in guinea pigs and humans (Ward et al., 1995), and inhaled NO exerts bronchodilatory effects against methacholine-induced bronchoconstriction in vivo (Kacmarek et al., 1996). Studies in bronchial and tracheal smooth muscle have shown that a major target of NO is the enzyme soluble guanylyl cyclase (Ellis, 1997), although NO is also a very active agonist at inducing vascular smooth muscle relaxation. There is therefore a need for suitable NO donors that have minimal effects on the vasculature.

NO donors can produce nitrites (NO₂⁻) and nitrates (NO₃⁻) and mediate their relaxing effects through the cGMP-mediated pathway. The main target of cGMP is the family of cGMP-dependent protein kinases. Activation of these kinases and subsequent phosphorylation of various proteins constitute a cascade of reactions that cause the reduction of cytosolic Ca²⁺ concentration, leading to relaxation (McDaniel et al., 1992). However, not all studies support the role of soluble guanylyl cyclase/cGMP pathway in mediating the relaxant effect of NO, and cGMP-independent pathways are increasingly being recognized (Redington, 2006). The controversial results of these studies may be due to the different species and NO donors used.

The main prototypes of NO donors traditionally used, such as sodium nitroprusside and nitroglycerine, have several well known adverse effects, such as rapid tachycardia, high toxicity, and rapid induction of tolerance. Likewise, sydnonimines, another well known class of NO donor drugs, have a characteristically low therapeutic index (because of cyanide toxicity) (Scatena et al., 2010).

At present, a number of groups are designing and synthesizing various chemical compounds capable of modulating NO metabolism for therapeutic purposes that also possess an improved therapeutic index (Scatena et al., 2010). Specifically, various new classes of NO donors are under intense pharmacological investigation (such as S-nitrosothiols, diazeniumdiolates, furoxans, and zeolites), each characterized by a particular PK and PD profile. The most important obstacle in the field of new NO donor drugs seems to be carefully targeting NO release to lungs at an optimal concentration to achieve a beneficial action and to limit possible adverse effects, particularly on the cardiovascular system.

α′-[[(1,1-Dimethylethy)amino]methyl]-4-hydroxy-1,3-benzenedimethanol nitrate (NCX-950), a NO-releasing albuterol, elicits potent relaxant and anti-inflammatory activities compared with albuterol (Lagente et al., 2004). These effects seem to be associated mainly with the stimulation of β₂-AR. However, it has been suggested that the release of NO from the NO moiety may be involved in the effects of NCX-950 because the effects of NCX-950, at equimolar doses, are slightly but significantly superior to that of albuterol alone (Lagente et al., 2004).

Budesonide 21-[(4′-nitrooxymethyl)benzoate] (TPI 1020, formerly NCX 1020 or NO-budesonide) is a novel compound, currently in development for the treatment of chronic respiratory disorders, with a dual mechanism of action involving corticosteroid activity (derived from its metabolite budesonide) and donation of NO. TPI 1020 has been shown to be superior to budesonide in an animal model of COPD, inhibiting airway hyper-responsiveness and reducing neutrophil infiltration into lungs (Nevin and Broadley, 2004). It has been shown that the NO component of TPI 1020 potentiates the bronchodilator activity of albuterol, and their combination lasted longer than either drug administered individually (Turner et al., 2010).

Another new NO donor TERPY (ruthenium complex [Ru(terpy)(bdq)NO⁺]³⁺) has been shown to be more effective in reducing cytosolic Ca²⁺ concentration and inducing relaxation in rat trachea greater than that of the sodium nitroprusside alone (Castro et al., 2011). In contrast to sodium nitroprusside, the NO released from the ruthenium complex induces ASM relaxation by cGMP-independent mechanisms and seems to involve Ca²⁺ and K⁺ fluxes across the membrane, leading to decreased cytosolic Ca²⁺ concentrations.

G. E-Prostanoid Receptor 4 Agonists

Inhaled prostaglandin E₂ (PGE₂) has been shown to be a bronchodilator in subjects with asthma (Melillo et al., 1994) or COPD (Kawakami et al., 1973). However, PGE₂ itself has the potential to cause the adverse effects of cough and retrosternal burning when inhaled by humans, particularly in subjects with asthma (Choudry et al., 1989). Moreover, PGE₂ also stimulates other prostaglandin-like receptors having inhibitory and stimulatory activity on the intrinsic tone of the airway smooth muscle. PGE₂ acts predominantly via specific E-prostanoid (EP) receptors. Whereas it has been established that PGE₂-induced sensory nerve activation and cough are mediated via the EP₃ receptor (Maher et al., 2009), the EP₄ receptor seems to be the predominant receptor responsible for PGE₂-induced relaxation of human ASM in vitro (Buckley et al., 2011). Stimulation of the EP₄ receptor with PGE₂ leads to a directly stimulation of the adenylyl cyclase via the Gα_s subunit, which converts ATP to cAMP. Thus, highly potent EP₄ subtype-selective receptor agonists have been suggested to have therapeutic potential without side effects (Buckley et al., 2011).

H. Bitter Taste Receptor Agonists

Bitter taste receptors work as chemoreceptors that interact with taste stimuli to initiate an afferent signal to the brain, where it becomes taste perception. Stimulation of the taste 2 receptors is responsible for the bitter taste (Behrens et al., 2007). These receptors were recently found on airway smooth muscle (Deshpande et al., 2010). When activated, they caused relaxation through a calcium-dependent mechanism. Agonists to these receptors may make up a new class of useful direct bronchodilators for treating obstructive lung disease, but because they are members of the G protein-coupled receptor superfamily, they may undergo desensitization (Robinett et al., 2011). Saccharin, chloroquine, denatonium, aristocholic acid, strychnine, quinine, colchicine, and yohimbine were used in the preclinical studies. Aristocholic acid, strychnine, and yohimbine are probably too toxic to be considered for human use. Quinine was selected for use in the mouse model of asthma, possibly because it was expected to be effective with minimal toxicity (Deshpande et al., 2010). In the low doses of quinine used in the treatment of muscle cramps, the incidence of serious adverse effects is not higher than that for the placebo (El-Tawil et al., 2010). Thus, perhaps it is worth considering developing an inhalation preparation of quinine for testing in humans with asthma (Doggrell, 2011).

A. Pharmacologic Rationale for Combination Therapy

B. Combination Therapy of β₂-Adrenergic Receptor Agonists with Muscarinic Acetylcholine Receptor Antagonists

Early studies questioned the value of combination therapy with ipratropium bromide and albuterol (Easton et al., 1986). However, several large trials found that concomitant administration of the two drugs by MDI (COMBIVENT Inhalation Aerosol Study Group, 1994) or nebulization (COMBIVENT Inhalation Aerosol Study Group, 1997) had additive bronchodilator effects.

The introduction of LABAs gave physicians additional therapeutic options for COPD. Results from two short-term studies that evaluated small COPD-patient cohorts suggested no substantial additive effect when combining salmeterol or formoterol with ipratropium bromide 40 μg (Matera et al., 1996; Sichletidis et al., 1999). In contrast to these earlier studies, van Noord et al. (2000) demonstrated that a 12-week treatment with 50 μg of salmeterol b.i.d. plus 40 μg of ipratropium bromide four times daily was more effective than the salmeterol alone in improving FEV₁ and specific airway conductance. Subsequently, D'Urzo et al. (2001) demonstrated that adding formoterol (12 μg b.i.d.) to ipratropium bromide (40 μg four times a day) was more effective than the addition of albuterol (200 μg four times a day) in patients with COPD who required combined bronchodilator therapy.

Addition of LABAs to tiotropium bromide has also been studied in short-term trials (≤6 weeks) (Cazzola et al., 2004a,b; van Noord et al., 2005, 2006). Taken together, these studies indicate that once-daily formoterol/tiotropium bromide and salmeterol/tiotropium bromide combinations provide significant additive effects in patients with moderate to severe COPD and are well tolerated. A bronchodilator-mediated symptom benefit of the once-daily combination is also reflected in the significant decrease in albuterol use as rescue therapy and symptoms of dyspnea. However, additional and more consistent improvements during the day and night were provided when the LABA was dosed twice daily.

Several other studies examined the regular treatment with LABA plus tiotropium bromide combination. In a 12-week study comparing 12 μg of formoterol b.i.d. plus 18 μg of tiotropium bromide once daily with tiotropium bromide monotherapy, the combination significantly increased the FEV₁ area under the curve at 0 to 4 h compared with tiotropium bromide at all time points (Tashkin et al., 2009). At the end of the study, increases from baseline in trough FEV₁ and FVC were significantly greater with the combination treatment than with tiotropium bromide alone. These outcomes were again found in a longer study lasting 6 months using similar treatment arms (Vogelmeier et al., 2008). This study also showed that the combination treatment arm had fewer exacerbations, although the difference between arms did not reach statistical significance. Finally, another study comparing salmeterol plus tiotropium bromide versus tiotropium bromide alone found that the combination improved quality of life scores, but not FEV₁, and was unable to decrease the number of exacerbations over tiotropium bromide alone (Aaron et al., 2007). Overall, the outcomes of these studies demonstrate that there are a number of added benefits in using combinations over tiotropium bromide alone. LABA/LAMA combination seems to play an important role in maximizing bronchodilation, studies to date indicating that combining different classes of bronchodilators results in significantly greater improvements in lung function and other outcomes compared with individual drugs used alone and that these combinations are well tolerated in patients with moderate to severe COPD (Cazzola and Tashkin, 2009).

C. Novel Long-Acting β₂-Adrenergic Receptor Agonists (Ultra-Long-Acting β₂-Adrenergic Receptor Agonists) and Long-Acting Antimuscarinic Agent Combinations under Development

It has been documented that adding indacaterol to tiotropium bromide is a beneficial strategy for patients with moderate to severe COPD, affording greater bronchodilation and lung deflation (as reflected by the increase in IC) (Mahler et al., 2011). The need for rescue medication was decreased after treatment with such combinations. The safety and tolerability profiles were similar in both treatment groups. It is not surprising, therefore, that current opinion is that it will be advantageous to develop inhalers containing combinations of two or more classes of once-daily long-acting bronchodilator drugs, in an attempt to simplify treatment regimens as much as possible (Cazzola and Matera, 2008, 2009; Cazzola et al., 2011). Nonetheless, some new combinations, such as formoterol plus glycopyrronium bromide and formoterol plus aclidinium bromide (LAS-40464) are under development with an expected twice-daily dosing regimen (Matera et al., 2011a,b).

A range of once-daily LABA (ultra-LABAs) and LAMA fixed-dose combinations, including QVA149 (the combination of indacaterol and the LAMA glycopyrronium bromide), olodaterol plus tiotropium bromide, and vilanterol plus umeclidinium bromide, are in clinical development as fixed combinations (Cazzola and Matera, 2008, 2009; Cazzola et al., 2011). There is documentation that a 7-day treatment with QVA149 (300 μg of indacaterol/50 μg of glycopyrronium bromide) once daily via a single-dose dry powder inhaler is more effective than 300 and 600 μg of indacaterol (van Noord et al., 2010). Moreover, QVA149 at the dosage of 600/100, 300/100, or 150/100 μg has a safe cardiovascular profile, with no different 24-h mean heart rate at day 14 in QVA149 and placebo or indacaterol, and no clinically relevant differences in QTc intervals observed among treatment groups on days 1, 7, and 14 (Van de Maele et al., 2010). Addition of olodaterol enhanced the beneficial effect of the tiotropium bromide monotherapy on ACh-induced bronchoconstriction with a longer duration of action in anesthetized dogs (Bouyssou et al., 2010c). Moreover, the protective effect of olodaterol against histamine-induced bronchoconstriction in guinea pigs was augmented and prolonged in the presence of tiotropium bromide, which had no significant effect by itself (Meurs et al., 2011). Olodaterol and tiotropium bromide used in combination might control some proinflammatory events in COPD. In fact, they synergistically control TGFβ1-mediated neutrophilic inflammation, at least in vitro (Profita et al., 2011). It has also been shown that olodaterol/tiotropium bromide (10/5 μg) fixed-dose combination administered using the Respimat soft mist inhaler is more effective than 5 μg of tiotropium bromide in patients with COPD, with superior bronchodilation over 24 h after 4 weeks of once-daily dosing (Maltais et al., 2010).

D. Bifunctional Muscarinic Acetylcholine Receptor Antagonists and β₂-Adrenergic Receptor Agonists

An alternative strategy is the design and development of dual-acting agents that combine both mAChR antagonist and β₂-AR agonist pharmacology in a single molecule (Cazzola and Matera, 2009; Cazzola et al., 2011) by conjugating a β₂-AR agonist motif onto that of an M₃ mAChR antagonist. This would create larger molecules with high lipophilicity that are likely to possess high metabolism and low oral absorption to minimize potential systemically driven side effects from the swallowed component after inhalation (Jones et al., 2011b). This approach may offer several advantages over combination therapy of two separate drug entities. These include the benefit of delivering a fixed ratio into every region of the lung, reducing the complexity of combination inhalers (Cazzola and Matera, 2009; Cazzola et al., 2011); a single PK profile; a uniform ratio of activities at the cellular level; and a simplified clinical development program (Steinfeld et al., 2011). These agents are known as dual-acting mAChR antagonist–β₂-AR agonist (MABA) bronchodilators. All MABA compounds disclosed so far have a M₃ mAChR antagonist moiety and a β₂-AR agonist moiety connected by a linker, usually an aliphatic and linear linker with 7 to 10 atoms. Such compounds may feature the diphenylcarbamate moiety, a tiotropium-like moiety, or the tolterodine moiety. However, it must be mentioned that although the marketed inhaled mAChR antagonists ipratropium bromide and tiotropium bromide both contain quaternary ammonium groups, the incorporation of a quaternary ammonium pharmacophore into a bifunctional MABA constructs results in lower affinities for the M₃ mAChR (Hughes et al., 2011).

TEI3252 (structure not disclosed) is a novel bifunctional bronchodilator that in experimental settings showed bronchoprotective activities against ACh- and histamine-induced contraction in a dose-dependent manner at the dose range of 1 to 5 μg/kg (Sugiyama et al., 2010). The efficacy of TEI3252 was long-lasting compared with existing bronchodilators, such as tiotropium bromide and indacaterol. On the other hand, the inhibitory effect on salivation was not observed even at the dose up to 100 μg/kg. This finding suggests that TEI3252 has a reduced side-effect profile.

Evaluation of THRX-200495 (structure not disclosed), a single bifunctional molecule that possesses both mAChR antagonist and β₂-AR agonist pharmacology, in a guinea pig model of bronchoconstriction revealed a matched mAChR antagonist and β₂-AR agonist potency [median dose that causes 50% inhibition (ID₅₀) = 11.4 and 11.2 μg/ml, respectively], with similar onset of action and potent dual pharmacology (MABA ID₅₀ = 3.5 μg/ml) lasting for >24 h (McNamara et al., 2009).

[1-[9[[(2R)-2-Hydroxy-2-(8-hydroxy-2-oxo-1H-quinolin-5-yl)ethyl]amino]nonyl]piperidin-4-yl] N-(2-phenylphenyl)carbamate (GSK961081, formerly TD5959) is a further novel bifunctional molecule. It conferred potent 24-h bronchoprotection in guinea pigs through a dual mechanism involving antagonism of mAChRs and agonism of β₂-ARs. Dual pharmacology yielded bronchoprotection that was 2- to 5-fold more potent (MABA ID₅₀ = 6.4 mg/ml) than either ipratropium bromide or albuterol alone (Pulido-Rios et al., 2009). In phase I randomized double-blind placebo-controlled single- and multiple-dose studies that enrolled healthy volunteers, GSK961081 was generally well tolerated and demonstrated evidence of bronchodilation over 24 h after a single dose and after seven consecutive daily doses; consequently, it has entered into phase II (Cazzola and Matera, 2009). In a phase II study, GSK961081 dosed at both 400 and 1200 μg once daily showed bronchoprotection at day 14 that was at least equivalent to that of 50 μg of salmeterol b.i.d. plus 18 μg of tiotropium bromide once daily, as measured by changes in FEV₁ (Cazzola and Matera, 2009). Both the time to peak effect and maximum bronchodilation of GSK961081 were numerically better than salmeterol plus tiotropium bromide, although the study was not powered to compare the results to the salmeterol plus tiotropium bromide control. In patients with COPD, after 14 days' dosing, 400 and 1200 μg of GSK961081 induced a significant improvement in trough FEV₁ (0.115 and 0.168 liters, respectively) similar to that elicited by salmeterol plus tiotropium bromide (0.103 liters) (Norris et al., 2011). There was no significant difference in maximum change from baseline heart rate or QTc after the final dose of any active treatment versus placebo. Adverse events were similar across all groups with the exception of tremor (n = 2, 1200-μg dose) and dry mouth (n = 1, 1200-μg dose), seen after GSK961081 only.

PF-3429281 (structure not disclosed) is another inhaled dual mAChR antagonist/β₂-AR agonist. In anesthetized guinea pigs, it caused a dose-related inhibition of ACh-induced bronchoconstriction (Philip et al., 2010). This was 26-fold weaker than tiotropium bromide and equipotent with salmeterol. Infusion of propranolol throughout the experiment blocked the effects of salmeterol on ACh-induced bronchoconstriction, whereas both tiotropium bromide and PF-3429281 were unaffected. Data generated in vitro using guinea pig isolated trachea suggest that the duration of action of the β₂-AR component is longer than the M₃ mAChR component (Patel et al., 2010a). In an anesthetized dog model of bronchoconstriction, PF-3429281 had an potency equivalent to that of ipratropium bromide and a superior therapeutic index and duration of action compared with salmeterol (Wright et al., 2010).

PF-4348235 (structure not disclosed) is a novel and potent MABA (Patel et al., 2011). PF-4348235 K_i at the M₃ mAChR was 0.79 nM (0.52–1.21 nM). This was ∼80-fold less potent than tiotropium bromide; however, both exhibited similar mAChR selectivity. PF-4348235 gave an apparent t_{1/2 off} of 117.1 min (95% CI, 64.2–213.6 min) at the M₃ mAChR. This was significantly longer than ipratropium bromide at t_{1/2 off} of 11.8 min (95% CI, 10.2–13.8 min) and shorter than tiotropium bromide at t_{1/2 off} >1370 min (95% CI, 1310–1420 min). PF-4348235 potency was 3.7 nM (95% CI, 2.2–6.2 nM) at the human β₂-AR, which was ∼6-fold less potent than salmeterol. After washing, PF-4348235 gave a fold shift to the right (RWS) of 1.5, which was not significantly different from that of salmeterol (2.2 RWS) but was significantly smaller than that of formoterol (97.2 RWS). It demonstrated >1000-fold selectivity over the human β₁-AR, similar to salmeterol and formoterol. PF-4348235 exhibited single and dual pharmacology in vivo using conscious guinea pigs with a potency and duration of action similar to salmeterol (Hulland et al., 2011).

E. β₂-Adrenergic Receptor Agonist or Long-Acting Antimuscarinic Agent/Phosphodiesterase Inhibitor Combinations

There is evidence that (R,R)-glycopyrrolate can have beneficial interactions with the PDE4 inhibitor rolipram in inhibiting inflammatory mediators (Pahl et al., 2006).

In patients with moderate to severe COPD treated with salmeterol or tiotropium bromide, roflumilast improved lung function and some clinically relevant symptomatic outcomes (Fabbri et al., 2009). The improvement in pre- and postbronchodilator FEV₁ means that the beneficial effect of roflumilast on lung function is in addition to that achieved with standard bronchodilators, an effect that is probably not due primarily to smooth muscle relaxation but to other mechanisms. It has been suggested that the improvement in lung function might be associated with a reduction in numbers of sputum neutrophils and eosinophils in patients with COPD. Apparently, the addition of roflumilast to tiotropium bromide induces a larger improvement in prebronchodilator FEV₁ than its addition to salmeterol, and this was not an unexpected finding considering the mechanism of actions of the two drugs (Cazzola et al., 2010c). However, it must also be mentioned that some trials documented that the addition of a LABA did not change the mean prebronchodilator FEV₁ increase observed with roflumilast in monotherapy (Calverley et al., 2009).

F. Combination Therapy of β₂-Adrenergic Receptor Agonists with Inhaled Corticosteroids

Combination therapy of ICS and inhaled LABA has become the “gold standard” therapy for asthma management. In adolescent and adult patients with asthma with suboptimal control on low-dose ICS monotherapy, the combination of LABA and ICS is modestly more effective in reducing the risk of exacerbations requiring oral corticosteroids than a higher dose of ICS. Combination therapy also led to modestly greater improvement in lung function, symptoms, and use of rescue β₂-AR agonists and to fewer withdrawals as a result of poor asthma control than with a higher dose of ICSs (Ducharme et al., 2010a). In asthmatic adults who are symptomatic on low to high doses of ICS monotherapy, the addition of a LABA at licensed doses reduces the rate of exacerbations requiring oral steroids, improves lung function and symptoms, and modestly decreases use of rescue short-acting β₂-AR agonists (Ducharme et al., 2010b). In children, the effects of this treatment option are much more uncertain (Ducharme et al., 2010b). LABAs should not be used as monotherapy in patients with asthma, but rather only with an ICS; however, because of issues in using two inhalers for two different drugs, combination inhalers are now considered standard (Sears, 2011).

The LABA/ICS combination is now increasingly recommended in patients with COPD because, in general, the addition of LABA to ICS provides additional benefits (Cazzola and Hanania, 2006; Hanania, 2008). Several large-scale studies in patients with moderate to severe COPD have demonstrated that treatment with salmeterol/fluticasone propionate and formoterol/budesonide leads to significantly greater improvements in lung function, exacerbations, health status, and breathlessness compared with placebo or monotherapy with the component drugs, although combination treatments may differ with regard to specific outcomes. Compared with placebo, combination therapy leads to a significant reduction of a quarter in exacerbation rates (Nannini et al., 2007c). Compared with mono component ICS therapy, LABA/ICS combination significantly reduces morbidity and mortality in COPD (Nannini et al., 2007a). Moreover, combination therapy is more effective than LABAs in reducing exacerbation rates, although the evidence for the effects on hospitalizations is mixed (Nannini et al., 2007b). There is a significant reduction in all-cause mortality with the addition of data from the Toward a Revolution in COPD Health (TORCH) study, although in the TORCH study regular treatment with salmeterol/fluticasone propionate narrowly missed demonstrating a statistically significant benefit on the reduction in all-cause mortality over 3 years (17.5% reduction in risk, P = 0.052) (Calverley et al., 2007). In addition, in the TORCH study, the LABA/ICS combination reduced the rate of decline in FEV₁ in patients with moderate to severe COPD by 16 ml/year compared with placebo. However, this improvement was also observed in the LABA-only and ICS-only groups (Celli et al., 2008). Moreover, the superiority of combination inhalers should be viewed against the increased risk of side effects, particularly pneumonia in patients with COPD (Nannini et al., 2007c).

G. Combination Therapy of Long-Acting Muscarinic Agonists with Inhaled Corticosteroids

Very few studies published to date have been designed specifically to evaluate the effect of ICS and tiotropium bromide combinations on clinical outcomes, and this is an area that warrants future study. A retrospective analysis of the pooled results of two similar 6-month, randomized, double-blind, double-dummy, placebo-controlled parallel group studies of inhaled tiotropium bromide (18 μg once daily) and salmeterol (50 μg b.i.d.) in the treatment of patients with COPD demonstrated superior spirometric responses with tiotropium bromide plus ICS compared with salmeterol plus ICS (Hodder et al., 2007). In addition, clinically important endpoints such as dyspnea, health status, and frequency of exacerbations were superior with tiotropium bromide plus ICS compared with salmeterol plus ICS (Hodder et al., 2007).

Treatment with tiotropium bromide and budesonide in patients with COPD has led to significant improvements in the quality of life status according to SGRQ (Um et al., 2007), measuring both dyspnea and lung function, and a reduction of the number of exacerbations over 6 months of treatment in patients with COPD and chronic asthma (Choi et al., 2007). It is noteworthy that the effects of tiotropium bromide added to ICSs have also been explored in patients with asthma, demonstrating noninferiority to the concomitant treatment with LABAs and ICSs (Lee et al., 2008). These results suggest the possibility of LAMA as an alternative to LABA for the treatment of asthma.

H. Other Possible Combination Therapies

A number of studies have reported that when theophylline and β₂-AR agonists are administered together (Cazzola et al., 2000; ZuWallack et al., 2001), some added clinical benefit is gained compared with the drugs administered alone. However, it has been questioned whether this has been due to the use of a suboptimal dose of inhaled β₂-AR agonist in the studies. In any case, it has been shown that regular theophylline treatment neither prevents nor worsens the development of tolerance to the bronchoprotective effect of salmeterol (Cheung et al., 1998).

The addition of theophylline to ICSs therapy is clinically beneficial in patients with asthma or COPD. It has been shown that combined treatment with low-dose inhaled budesonide plus low-dose theophylline yields similar improvement in lung function tests compared with patients receiving high-dose budesonide alone, without the adrenal suppression associated with high dose-budesonide (Evans et al., 1997). Nonetheless, a recent meta-analysis of randomized clinical trials has suggested that the addition of theophylline elicits a greater increase in lung function than increasing the dose of ICS, but only in short-duration treatment (Wang et al., 2011). However, these results should be interpreted with caution because of the small sample of included studies and the possible presence of a publication bias (Wang et al., 2011).

Theophylline has been shown to activate HDAC, resulting in a marked potentiation of the anti-inflammatory effects of ICSs (Ito et al., 2002). In patients with COPD, the combination of theophylline and ICS is more effective in reducing airway inflammation than either drug alone (Ford et al., 2010c). This is now leading to therapeutic trials in COPD with low doses of theophylline. Low-dose theophylline also improves asthma control in asthmatic patients who smoke and show no response to ICS alone (Spears et al., 2009).

I. Triple Combination Therapies

There is limited documented clinical evidence for the use of triple therapy in COPD, but studies published to date indicate that ICS/LABA in combination with tiotropium bromide improves lung function, COPD symptoms, and health status and reduces the risk of hospitalizations compared with tiotropium bromide alone in patients with moderate to severe COPD (Aaron et al., 2007; Cazzola et al., 2007; Singh et al., 2008a; Welte et al., 2009). Intriguingly, a proof-of-concept study has recently shown that the addition of once-daily tiotropium bromide to a high-dose ICS plus a LABA significantly improves lung function over 24 h in patients with inadequately controlled severe, persistent asthma (Kerstjens et al., 2011). In any case, Bouyssou et al. (2011) has shown that olodaterol in combination with tiotropium bromide and ciclesonide elicits potent synergistic bronchoprotective activity in the guinea pig. This effect is significantly higher than the summarized values of the respective monotherapies and supports the concept of the triple therapy.

Clearly, a triple inhaler would be advantageous from an adherence point of view because of the convenience of having three inhalation products in one inhaler. This improved patient compliance and adherence to the therapy regimen will certainly affect the management of such a chronic condition but will also reduce the repeating error that arises from the incorrect use of two or more inhaler devices. However, there is no hard evidence to date that this approach is useful or even necessary (Salama et al., 2011).

It is likely that the development of once-daily dual-action ultra LABA + LAMA combination products may serve as a basis for improved “triple-therapy” combinations through coformulation with novel anti-inflammatory compounds such as inhaled PDE4 inhibitors that could deliver three complementary therapeutic effects for patients with COPD (Cazzola and Matera, 2006); in addition, the use of mixed PDE3/4 inhibitors such as RPL554 in combination with an M₃ mAChR antagonist may prove beneficial. In any case, the development of once-daily dual-action ultra LABA plus LAMA combination products may serve also as a basis for improved triple-therapy combinations through coformulation with novel ICSs. The potential for these therapeutic strategies to be administered once daily simplifies patient treatment regimens and therefore increases the likelihood of compliance with therapy.

The only inhalation triple formulation commercially available is an HFA pressurized MDI suspension that contains tiotropium bromide monohydrate equivalent to bromide 9 μg of tiotropium, 6 μg of formoterol, and 20 μg of ciclesonide. The three drugs are suspended in HFA 227, with apparently no other additives (Salama et al., 2011).

A combination of indacaterol, glycopyrronium bromide, and mometasone seems to be a real possibility (Fitzgerald and Fox, 2007). A combination of vilanterol with umeclidinium bromide and fluticasone furoate is another valid option. Furthermore, LAS-40369 (structure not disclosed), an association of aclidinium bromide and formoterol plus an undisclosed ICS under development (Matera et al., 2011b), is another combination under development. A combination that features the inhaled PDE4 inhibitor tofimilast with a LAMA (potentially tiotropium bromide) in addition to, potentially, a MABA seems to be less realistic (Fitzgerald and Fox, 2007). It is intriguing that a family of dual-action M₃ mAChR antagonists/PDE4 inhibitors has recently been discovered (Provins et al., 2006: Provins et al., 2007). The pharmacological profile of 6-(azepan-1-yl)-N,2-dicyclopropyl-5-methylpyrimidin-4-amine (UCB-101333-3), a 4,6-diaminopyrimidine, is intriguing (Provins et al., 2007). The addition of an ultra-LABA to UCB-101333-3 could create an interesting combination for treating asthma or COPD (Cazzola and Matera, 2009).

A. β₂-Adrenergic Receptor Agonists

B. Muscarinic Acetylcholine Receptor Antagonists

As already mentioned, concerns have been raised about possible associations of mAChR antagonists with cardiovascular morbidity and mortality, and a body of evidence supports the possible existence of a link between the use of SAMAs and cardiovascular risk. Two pooled analyses of randomized controlled trials (Singh et al., 2008b; Worth et al., 2011) and a retrospective, nested case-control study (Lee et al., 2008) reported increased risks for cardiovascular events or mortality with inhaled mAChR antagonists use. The highest risk for mortality was among those with pre-existing arrhythmias.

A number of studies have investigated the influence of mAChR antagonists on the cardiovascular system. In humans, the mean maximal heart rate rose significantly for 0.03 mg/kg i.v. atropine (+46.2%) and for 0.03 mg/kg i.v. ipratropium bromide (+57.4%) (Libersa et al., 1988). The effects obtained with ipratropium bromide on the heart rate lasted nearly twice as long as those obtained with atropine (120 ± 38.4 and 70 ± 30 min, respectively, for the pharmacological half-life). In 13 patients with sinus bradycardia, heart rate and hemodynamics were recorded before and 3, 30, 60, and 120 min after administration of 0.5 mg i.v. ipratropium bromide. Heart rate already started to rise during the injection of ipratropium bromide and increased by an average of 78% 3 min after administration compared with pretreatment control values (Kikis et al., 1982). Mean heart rate was still markedly increased by an average of 26% at 120 min after the injection of the drug. This effect of ipratropium bromide on heart rate was accompanied by an increase in cardiac index, whereas stroke volume decreased because of a decrease in ejection fraction and diastolic filling time.

Even so, acute inhalation of ipratropium bromide seemed to have no significant effect either on cardiac vagal tone or on heart rate, at least in young asthmatics (Lehrer et al., 1994). This is not surprising. In fact, although ipratropium bromide is structurally similar to atropine and has similar actions if given parenterally, it is a quaternary ammonium compound that has low oral bioactivity and few extrapulmonary effects, even after administration of very large doses.

The large UPLIFT trial has clarified the doubts generated by the reported retrospective analyses of using tiotropium bromide (Tashkin et al., 2008). In this study, cardiovascular adverse events were significantly less with tiotropium bromide (RR, 0.84; 95% CI, 0.73–0.98), although 74% of patients reported having received ICSs, 72% LABAs and 46% a fixed combination of the two (Tashkin et al., 2008). The RR of chronic heart failure (RR, 0.59; 95% CI, 0.37–0.96) and myocardial infarction (RR, 0.71; 95% CI, 0.52–0.99) was also significantly lower with tiotropium bromide compared with placebo, and those of angina (RR, 1.44; 95% CI, 0.91–2.24) and cardiac failure (RR, 1.25; 95% CI 0.84–1.87) were not significantly higher with tiotropium bromide compared with placebo. The RR of stroke in the tiotropium bromide versus placebo groups was not significantly different (RR, 0.95; 95% CI, 0.70–1.29).

Because patients receiving tiotropium bromide with the mist inhaler are, however, potentially exposed to higher concentrations and the powder and mist inhaler formulations are considered to be distinct products (US Food and Drug Administration, 2009), Singh et al. (2011b) systematically reviewed and performed a meta-analysis on data from randomized controlled trials on the risk of mortality associated with inhaled tiotropium bromide delivered by the mist inhaler compared with placebo in patients with COPD. This meta-analysis of 6522 subjects documented that tiotropium bromide administered by the mist inhaler was associated with a 52% increased risk of all-cause mortality compared with placebo. The annualized number needed to treat for mortality associated with tiotropium bromide was estimated at 124 (95% CI, 52–5682). This means that one excess death would be expected for every 124 patients treated with 5 μg of tiotropium bromide for 1 year.

However, it is noteworthy that none of the examined studies had cardiovascular outcomes as the primary outcome, and the data reported are based on adverse event reporting, Thus cardiovascular parameters may not have been prospectively defined in a uniform fashion across the trials included in this meta-analysis and, consequently, there may have been important discrepancies between the trials in reporting of these outcomes (Cazzola et al., 2010a). Also of note is the fact that patients with a recent history of myocardial infarction, arrhythmia, or heart failure were excluded from the UPLIFT trial (Tashkin et al., 2008) (and possibly other studies); thus, it is not known whether such patients may be at an increased risk of any drug-related cardiovascular events in a real-world setting when using tiotropium bromide.

A. Older People

Aging-related modifications in lung mechanics, in receptor populations, in nervous system control, etc., might be responsible for a different effectiveness of bronchodilators in older patients compared with younger subjects (Bellia et al., 2006). A modification of bronchodilator responses to β₂-AR agonists in older people has been suggested (Connolly et al., 1995). There is considerable evidence relating the reduction in β₂-AR affinity (or a reduced percentage of high-affinity receptors) with increasing age, possibly in association with receptor internalization in membrane-bound vesicles (Scarpace and Baresi, 1988). Abnormal postreceptor events may also be implicated in impaired β-adrenergic activity in older people.

In a study on bronchodilator responses to albuterol and ipratropium bromide, the FEV₁ response to the β₂-AR agonist declined with age; conversely, the response to ipratropium bromide was not related to age (Barros and Rees, 1990). The bronchodilator effects of albuterol after methacholine-induced bronchoconstriction was tested in a study carried out in healthy subjects (Connolly et al., 1995). This study showed that the older group (age range, 60–76 years) had a lower sensitivity to bronchodilator effects of albuterol, and this was interpreted by the authors as being due to an age-related decrease in airway β₂-AR responsiveness. However, a retrospective analysis, carried out in subjects with an FEV₁ reduced by at least 20%, documented that aging does not affect bronchodilator response to β-AR agonist after methacholine-induced bronchoconstriction (Parker, 2004). Furthermore, a study aimed at exploring the effect of age on bronchodilator responses in acute severe asthma concluded that age is not a predictor of response to β-AR agonists (Rodrigo and Rodrigo, 1997). However, in this study, the age groups were differentiated using the age of 35 years as cut-off level, which obviously was inadequate to allow inferences on older patients' (>65 years) response to the treatment.

β₂-AR agonists may have significant side effects, especially on the cardiovascular system in older people. For this reason, oral β₂-AR agonists are normally avoided in older patients, and inhalation is the route of choice. β₂-AR agonist-induced side effects are dose-dependent and include an increase in myocardial oxygen consumption, changes in blood pressure, arrhythmias, hypokalemia (Haalboom et al., 1985), nausea, and tremor. Some of these side effects may be increased by concomitant use of other drugs, thus hypokalemia can be aggravated by concomitant treatments promoting potassium loss, including diuretics, ICSs, and theophylline.

Relatively little is known about possible age-dependent alterations in human mAChRs, although it has been postulated that during the aging process in humans, number and functional responsiveness of lung M₂ mAChR decrease. Under normal conditions, M₂ mAChRs limit the release of ACh from parasympathetic nerves. Hence, loss of M₂ mAChRs results in increased ACh release and in increased vagal tone (Bellia et al., 2006). It has been suggested that older patients (>60 years) respond to a mAChR antagonist better than younger subjects who, conversely, benefit more from β₂-AR agonists (van Schayck et al., 1991). If this is true, older subjects may require increased doses of β₂-AR agonists, compared with mAChR antagonists, to obtain maximal bronchodilatory response. However, Kradjan et al. (1992) found no age-related differences in either the magnitude of response or in the time and dose required to reach peak effects after either albuterol or ipratropium bromide. This finding indicated that the patients were not less responsive to either drug compared with the young people. Nonetheless, a recent population-based, retrospective cohort study has shown that older adults initially prescribed LABAs for the management of moderate COPD seem to have lower mortality than those initially prescribed LAMA use (Gershon et al., 2011).

Older subjects are at greater risk of theophylline toxicity than younger patients, mainly because of a more frequent occurrence of concomitant diseases, such as liver dysfunction, cardiac failure, and fever, and for a more frequent use of polypharmacy, which may interfere with theophylline metabolism (Bellia et al., 2006). Therefore, when used in older patients, dosage adjustment of theophylline is mandatory because this drug has different PK in older people. It has been proposed that, in these subjects, serum concentrations of the drug should range between 8 and 12 mg/l (i.e., a lower level than the range proposed for the general population).

B. Pediatric Populations

The doses of β₂-AR agonist delivered with the use of an MDI with spacer and face mask for infants and small children are not yet well defined. Moreover, the risk of administering high doses of drug to small children is still unknown. It has been suggested that very young children may require higher relative doses because of possible problems in nebulizer technique or differences in PK (higher clearance) (Penna et al., 1993).

A direct comparison between four doses of inhaled formoterol (4.5, 9, 18, and 36 μg) with 50 μg of salmeterol in children found that although both medications increased FEV₁, the increase was significantly greater for doses of formoterol ≥9 μg compared with salmeterol (Pohunek et al., 2004).

The European Respiratory Society task force, which has reviewed the evidence for pediatric medicines in respiratory disease, has advised that as-needed SABAs are the first choice for rescue treatment of acute asthma in children (Smyth et al., 2010). Regular, daily use of SABAs may cause deterioration of asthma and is not recommended. However, the scientific basis for this issue in children is limited. Levalbuterol may have some advantages over racemic albuterol, but Qureshi et al. (2005) documented that there was no difference in clinical improvement in children with acute moderate to severe asthma exacerbations treated with either racemic albuterol or levalbuterol.

Single doses of LABAs provide at least 12 h of bronchodilation in asthmatic school-aged children. Bronchodilator and protective effects of single doses of LABA have been documented in children of preschool age and older. LABA for daily maintenance treatment in asthmatic children who remain symptomatic despite conventional doses of ICS is widely advocated and practiced but insufficiently supported by the literature and debated. In any case, the bronchodilator and bronchoprotective effects of LABAs in children become less during long-term use.

It must be mentioned that two Cochrane reviews have reported that the risk for serious adverse events with regular salmeterol or formoterol use compared with placebo is greater in children (Cates and Cates, 2008; Cates et al., 2009). A higher risk for serious adverse events associated with LABA therapy in children was also described by Rodrigo et al. (2009).

A small study that enrolled 48 steroid-naive atopic asthmatic children, 7 to 11 years of age, demonstrated that add-on therapy with montelukast to low dosage of budesonide is more effective than the addition of LABA or doubling the dose of budesonide in controlling airway inflammation, measured as the fraction of nitric oxide in exhaled breath (FE_NO), in asthmatic children (Miraglia del Giudice et al., 2007). Nonetheless, several meta-analyses of studies evaluating the safety and effectiveness of LABAs in adults and children with asthma published by the Cochrane Airways Group concluded that adding LABAs to ICSs reduces the requirement for rescue therapy, reduces exacerbations requiring systemic steroids, improves lung function compared with ICS alone, and is superior to adding an leukotriene receptor antagonist to ICS (Elkout et al., 2010).

Another Cochrane review (Plotnick and Ducharme, 2000) documented that a single dose of a mAChR antagonist is not effective for the treatment of mild and moderate exacerbations and is insufficient for the treatment of severe exacerbations. Nonetheless, adding multiple doses of a mAChR antagonist to a β₂-AR agonist seems safe, improves lung function, and would avoid hospital admission in 1 of 12 such treated patients. Although multiple doses should be preferred to single doses of mAChR antagonists, the available evidence supports their use only in school-aged children with severe asthma exacerbation. There is no conclusive evidence for using multiple doses of mAChR antagonists in children with mild or moderate exacerbations.

Children eliminate theophylline more rapidly on average than do adults and also show pronounced interindividual differences in the elimination of the drug (Ellis et al., 1976). Consequently, compared with adults, they tend to require relatively larger amounts of theophylline per day and the doses may have to be given at shorter intervals of time.

C. Patients with Heart Failure

Bronchodilator use is a powerful independent predictor of worsening heart failure and increased mortality in a broad spectrum of patients with heart failure. Whether this relates to a toxic effect of bronchodilators, underlying pulmonary disease, or both is unclear (Hawkins et al., 2010).

In particular, it remains unknown whether the use of β₂-AR agonists leads to an increased risk of heart failure and/or it may affect the risk of hospitalization among patients with existing chronic heart failure (CHF) (Matera et al., 2010). When acutely administered, β₂-AR agonists are efficacious in improving cardiac performance in patients with CHF because they enhance cardiac and stroke volume indexes in a dose-dependent manner (Matera et al., 2010). For these reasons, β₂-AR agonists are often used for the short-term enhancement of heart contractility and support of the circulation.

Nonetheless, data from the Acute Decompensated Heart Failure National Registry Emergency Module (ADHERE-EM) (Singer et al., 2008) demonstrated that 14% of patients who present with dyspnea are treated for COPD when COPD is absent and acute decompensated heart failure is the cause of dyspnea. Inhaled β₂-AR agonist use in patients with heart failure who do not have COPD seems to be associated with worse outcome. Au et al. (2003) identified an increased risk of heart failure hospitalization among patients with left ventricular (LV) systolic dysfunction who received a β₂-AR agonist within 3 months of the hospitalization. This risk increased as the number of canisters used per month increased; the effect persisted after adjustment for other pulmonary therapies, including ipratropium bromide and corticosteroids. The mortality risk was also higher for patients treated with a β₂-AR agonist, but the effect was significant only in patients who received three or more bronchodilator canisters per month.

There is evidence that β-AR agonists, although useful acutely, lead to increases in mortality with long-term use (Felker and O'Connor, 2001). Although β-AR agonists can relieve symptoms and improve hemodynamic parameters, their long-term administration has been associated with increased mortality in all sufficiently powered clinical trials with a variety of drugs ranging from strong to weak agonists. Actually, some reports suggest the presence of an association between β₂-AR agonists and the risk of incident CHF. The Washington DC Dilated Cardiomyopathy Study compared 129 case subjects with newly diagnosed idiopathic dilated cardiomyopathy with 258 randomly dialed neighborhood control subjects (Coughlin et al., 1995). An association between idiopathic dilated cardiomyopathy and history of emphysema or chronic bronchitis (OR, 4.4), asthma (OR, 1.9), oral β-AR agonists (OR, 3.4), and β₂-AR agonist inhalers or nebulization (OR, 3.2) was demonstrated. A total of 20% of the case subjects had a reported history of β₂-AR agonist inhaler use compared with 6.7% of the control subjects. On the contrary, the Ambulatory Care Quality Improvement Project (ACQUIP) (Au et al., 2004) found no association between the use of inhaled β₂-AR agonists and the risk of heart failure (1–2 canisters per month, OR, 1.3; ≥3 canisters per month, OR, 1.1). However, among the cohort with a history of CHF, there was a dose-response association between the number of inhaled β₂-AR agonists and the risk of hospitalization for CHF (1–2 canisters per month, adjusted OR, 1.8; ≥3 canisters per month, adjusted OR, 2.1). The increase in risk among patients with existing CHF was independent of history of COPD, corticosteroid use, β-AR blocker use, ACE inhibitor use, myocardial ischemia, and cardiovascular risk factors.

Whatever the case may be, the observation that β₂-AR agonists may exacerbate heart failure is supported by physiological observations (Matera et al., 2010). β₁-ARs are down-regulated and desensitized among patients with LV systolic dysfunction, and β₂-ARs, although desensitized, are found in normal numbers and represent a higher proportion of total β-ARs. Among patients with heart failure, β₂-AR agonists augment cardiac function, but, with regular exposure to β₂-AR agonists, myocardial β₂-ARs become desensitized and down-regulated. Moreover, long-term β₂-AR stimulation elevates G_i expression. The coupling of β₂-ARs to G_i proteins negatively regulates the G_s-mediated contractile response in the heart of many mammalian species. Intriguingly, the Ile¹⁶⁴ polymorphism of the β₂-AR, which leads to a dysfunctional receptor, has been demonstrated to be associated with decreased exercise tolerance and a 5-fold increased risk of death among patients with CHF. The possibility that long-term β-AR stimulation induces myocardial, but not systemic, elaboration of tumor necrosis factor-α, IL-1β, and IL-6 is another important finding. In fact, evidence suggests that proinflammatory cytokines are capable of modulating cardiovascular function by a variety of mechanisms, including promotion of LV remodelling, induction of contractile dysfunction, and uncoupling of myocardial β₂-ARs.

The possibility that β₂-AR agonists might exacerbate heart failure has led Ng et al. (2002) to explore whether long-term inhaled salmeterol therapy (100 μg b.i.d.) improved pulmonary function without augmentation of neurohormonal systems or ventricular ectopy in subjects with symptomatic heart failure and LV ejection fraction <40%. Salmeterol significantly increased mean rate-pressure product by 5%, and FEV₁ without producing measurable effects on neuroactivation or ventricular ectopy. The safe profile of salmeterol was also confirmed by the Drug and Safety Research Unit and the School of Medicine at Southampton (UK) (Jenne, 1998). Using the prescription-event monitoring technique, researchers from this Unit re-examined the safety issue for salmeterol and bambuterol, a long-acting oral agent and prodrug of terbutaline. There was an excess of nonfatal “cardiac failure” in the bambuterol group during the first month and a lower but increased incidence during the second to sixth months, whereas there was no excess in the salmeterol group. This finding raised the question of why bambuterol and not salmeterol caused CHF. It is likely that this was linked to the route of administration (oral instead of inhaled) and the different dose (significantly higher with bambuterol), which induced a high systemic bioavailability and, consequently, higher risk of side effects.

Although elevated sympathetic activity is associated with an adverse prognosis, a high level of parasympathetic activation confers cardioprotection by several potential mechanisms (Olshansky et al., 2008). These parasympathetic actions on the heart are mediated not only by the direct consequences of cardiac mAChR stimulation but also by a multitude of indirect mechanisms. Parasympathetic withdrawal, in addition to the well known augmentation of sympathetic drive, is an integral component of the autonomic imbalance characteristic of CHF (Binkley et al., 1991). Nonetheless, there is documentation that, in the setting of heart failure and sympathetic activation, muscarinic receptor stimulation decreases cardiac norepinephrine spillover, an effect not observed in patients with preserved LV function. (Azevedo and Parker, 1999). A recent cohort study that examined the association between mAChR antagonists use and cardiovascular events in a cohort of 82,717 U.S. veterans with a new diagnosis of COPD between 1999 and 2002 found an increased risk of cardiovascular events, mainly heart failure, associated with the use of ipratropium bromide within the preceding 6 months, whereas the risk was not present in subjects who received anticholinergics more than 6 months prior (Ogale et al., 2010). However, the large UPLIFT trial documented that the relative risk of congestive heart failure (RR, 0.59; 95% CI, 0.37–0.96) was significantly lower with tiotropium bromide compared with placebo (Tashkin et al., 2008).

D. Polypharmacy

In clinical practice, the use of multiple medications or polypharmacy for patients with COPD or asthma is commonly observed, and the potential for adverse effects and drug interactions is therefore increased (McLean and Le Couteur, 2004). Patients at highest risk include older people and those with multiple comorbid medical conditions. Adverse reactions related to polypharmacy and comorbidity are frequent, mainly in older patients.

Combining thiazide and loop diuretics with β₂-AR agonists may enhance hypokalemia and risk ECG changes, especially if the dosage of β₂-AR agonists exceeds the recommended range (Lipworth et al., 1989; Newnham et al., 1991). Concomitant use of formoterol and other β₂-AR agonists with drugs that prolong the QTc interval may potentiate the risk of ventricular arrhythmias, including torsades de pointes (Schering Corp., 2010). The risk of QT prolongation may be increased by cardiovascular disease, hypokalemia, hypomagnesemia, bradycardia, or genetic predisposition (e.g., congenital QT prolongation). The use of monoamine oxidase inhibitors and tricyclic antidepressants may also potentiate the vascular system effects of β₂-AR agonists, even when the latter were initiated within 2 weeks of discontinuation. Concomitant administration of oral roflumilast and inhaled formoterol under steady-state conditions does not affect the PD or PK of either drug (de Mey et al., 2011). In particular, there is no evidence of a relevant PD interaction with regard to myocardial repolarization or cardiac function in general. Moreover, the concomitant administration of roflumilast and formoterol did not negatively influence the safety profile of either drug. Although cardioselective β-AR blockers have been designed to target β₁-ARs but avoid β₂-ARs in the lung and elsewhere, so called cardioselective β-AR blockers (such as atenolol and bisoprolol) are only relatively selective and exert significant β₂-AR antagonism at therapeutic doses, although to a lesser extent than nonselective β-AR blockers such as propranolol. Thus, it might be considered counterintuitive to prescribe both β-AR blockers and β agonists in the same patient, even when they are targeting different organs (Cazzola et al., 2002c; Matera et al., 2010). In fact, there is some risk that bronchospasm may occur in certain patients and that the bronchodilator response to inhaled β₂-AR agonists might be impaired with these agents. Nonetheless, a recent retrospective cohort study documented that the addition of a β-AR blocker had no deleterious impact when added to a regimen that included a LABA (Short et al., 2011). Coadministration of albuterol and digoxin has shown to decrease digoxin serum levels (Edner and Jogestrand, 1989); therefore, close monitoring of the digoxin level in that context would be prudent.

As noted in section III.B, theophylline is metabolized by CYP1A2 and to a lesser extent by CYP3A3 and -2E1; medications that are inducers or inhibitors of these enzymes can affect the metabolism and elimination of theophylline. Consequently, a wide variety of drugs is known to alter the clearance of serum theophylline that may result in subtherapeutic effects or chronic intoxication (Aronson, 2006). Cardiac medications such as the β-AR blocker propranolol, the calcium channel blockers nifedipine and verapamil, and mexiletine reduce theophylline clearance by unknown mechanisms. Considering that increased theophylline concentrations can lead to an increased propensity for central and cardiac stimulation, it is prudent to avoid using a combination of theophylline and cardiac medications, especially antidysrhythmic drugs (Aronson, 2006). Combined therapy with theophylline and a β₂-AR agonist in young, otherwise healthy asthmatic subjects does not lead to an increase in the total number of ectopic beats but may increase the degree of complexity of ventricular premature beats (Coleman et al., 1986).

E. Use of Bronchodilation in Athletes and the Abuse of Bronchodilators

There is evidence that β₂-AR agonists can have an anabolic effect on muscle when given orally, but they are not ergogenic when given at the usual inhaled dose (McKenzie et al., 2002). These effects have been studied particularly for clenbuterol, but there is also proof that oral albuterol can improve performance because it increases strength, endurance, and power, although side effects can be unpleasant (Fitch, 2006). The ergogenic effects are associated with an increase in skeletal muscle protein, largely because of the inhibition of protein degradation (Maltin et al., 1993). In addition to the increase in muscle mass, β₂-AR agonists also decrease body fat (Yang and McElligott, 1989), and for this reason, they are also defined as “repartitioning agents.” The increase in muscle mass is associated with an increase in muscle force production (Dodd et al., 1996). The β₂-AR agonist-induced skeletal muscle hypertrophy is evident in fast- (type II) and slow- (type I) twitch muscles (Dodd et al., 1996), although in some muscles, a change in the proportion of fast- to slow-twitch fibers may occur (Burniston et al., 2007).

However, no effect has been reported in humans on oxygen uptake, lung function and performance with the use of therapeutic doses of inhaled bronchodilator in either patients with asthma or those without (Backer et al., 2007), but some studies have found that high doses of inhaled (van Baak et al., 2004) or systemic (Collomp et al., 2000) β₂-AR agonists have an ergogenic effect in patients without asthma. Although no effect on performance has been reported, increased lung function has been found in nonasthmatic elite athletes with inhaled β₂-AR agonists (Sue-Chu et al., 1999). Furthermore, some studies have noted a shorter period of breathlessness after exercise when healthy subjects have taken a β₂-AR agonist (Carlsen et al., 2001), indicating that the normal period with breathlessness after exercise is shortened when healthy subjects use a β₂-AR agonist.

Systemic use of β₂-AR agonists in an animal model increases weight and muscle mass. It is noteworthy that formoterol, and to a lesser extent salmeterol, are capable of producing skeletal muscle hypertrophy at microgram doses (Ryall et al., 2006). Nonetheless, inhaled formoterol neither improved endurance performance in healthy nonasthmatic athletes at hypobaric conditions equal to 2000 m above sea level (Riiser et al., 2006) nor improved endurance performance in cold environments compared with placebo (Tjørhom et al., 2007).

Inhaled β₂-AR agonists are the most effective bronchodilators for the relief of asthma symptoms and for pretreatment of exercise-induce asthma (EIA). However, daily use of SABAs may increase the severity of EIA (Hancox et al., 2002), daily administration of LABAs decreases the duration of their protection against EIA (Ramage et al., 1994), and the recovery from EIA after a bronchodilator is slower (Storms et al., 2004). Ideally, athletes should use β₂-AR agonists infrequently, but this may not be appropriate for those who train every day. Fortunately, EIA should be better controlled by use of other therapies. Such therapies are likely to target the production, release, and effects of the mediators of bronchoconstriction. Therefore, β₂-AR agonists should be reserved for occasional use and breakthrough symptoms (Fitch et al., 2008).

The management of the athlete with asthma should follow current national or international guidelines (e.g., Global Initiative for Asthma). Currently, there is no evidence that management of asthma in athletes should differ from that in nonathletes (Fitch et al., 2008). The International Olympic Commission (IOC) set up restrictions for the use of inhaled β₂-AR agonists in 1993 that have been modified repeatedly. The World Anti-Doping Agency (WADA) states that the use of β₂-AR agonists requires therapeutic use exemption and declared the following: “It is preferred to leave to the professional judgement of the physician the medical conditions under which these drugs are to be prescribed.” The team manager and team doctor are also responsible when an athlete is caught in doping. A concentration of urinary albuterol ≥1000 ng/l is considered an adverse analytical finding regardless of the grant of any form of therapeutic use exemption (Carlsen et al., 2008), bearing in mind that, in any case, after inhalation of doses up to 800 μg, urinary concentrations of albuterol are well below the limits used in doping control (Sporer et al., 2008).

There is evidence that mAChR antagonists may be effective against EIA in some patients, but not usually in the majority (Boulet et al., 1989). An additional protective bronchodilator effect may be obtained when ipratropium bromide is added to an inhaled β₂-AR agonist (Greenough et al., 1993). Freeman et al. (1992) reported that ipratropium bromide improved lung function both before and after exercise in asthmatic men compared with nonasthmatic men but had no effect on cardiorespiratory or cardiovascular parameters of performance after a step-wise cycle exercise test.

No ergogenic effects were attributable to theophylline therapy, which should therefore remain an acceptable means of management of athletes with asthma (Morton et al., 1989). However, a recent meta-analysis has documented that caffeine, which is a member of the methylxanthine group, significantly increases time to exhaustion compared with the placebo, and no dose-response relationship is evident for the effects of caffeine on time to exhaustion (Marshall, 2010).

F. Pregnancy

Witter et al. (2009) have suggested that the use of β₂-AR agonists increases the frequency of neurodevelopmental disorders in offspring, possibly resulting from β₂-AR agonist effects on the balance of sympathetic-parasympathetic tone, which may be particularly affected at specific developmental stages. However, based on the National Asthma Education and Prevention Program (NAEPP) review of six published studies that included a total of 6667 pregnant women, of whom 1929 had asthma and 1599 had taken β₂-AR agonists during pregnancy, data are reassuring regarding the safety of β₂-AR agonists used in pregnancy (National Heart, Lung, and Blood Institute National Asthma Education and Prevention Program Asthma and Pregnancy Working Group, 2005). Another large prospective study of 1828 women with asthma confirmed the lack of relationship between the use of inhaled β-AR agonists and adverse maternal or fetal outcome (Schatz et al., 2004).

There are no solid data in the literature indicating that one specific β₂-AR agonist should be considered preferable. Nevertheless, albuterol, terbutaline, and metaproterenol are considered SABAs of first choice, because they have been used for decades without any reported significant side effect in humans (Liccardi et al., 2003). Oral/parenteral administration of β₂-AR agonists to asthmatic pregnant women is not recommended for some important causes, such as the lack of safety data in the first trimester, the potential inhibitory effect on the delivery, and the higher rate of side effects (tachycardia and/or tremors) compared with the inhaled route (Liccardi et al., 2003).

Only few data are available on the safety during pregnancy of salmeterol and formoterol. Animal studies with salmeterol administered parenterally are not reassuring (Physician's Desk Reference, 1998). Thorough fetal examinations revealed no significant effects of maternal treatment with salmeterol in the rat, although ossification of the second or fifth sternebrae, hyoid bone, and eleventh thoracic centrum was marginally retarded at the highest dose of 2 mg/kg per day (Owen et al., 2010). Dutch rabbit fetuses exposed in utero exhibited significant effects at oral doses of 1 mg/kg per day or more (Owen et al., 2010). The effects observed included premature opening of the eyelids, cleft palate, sternebral fusion, and limb and paw flexures and a reduction in the ossification rate of the frontal cranial bones. Increased incidences of minor skeletal variants were also noted. However, the experience with chemically similar drugs such as albuterol (animal but not human studies reporting adverse effects) suggests that animal studies cannot be entirely transposed to humans (American College of Obstetricians and Gynecologists, and American College of Allergy, Asthma and Immunology, 2000). A postmarketing surveillance study of formoterol in general practice in England showed no important side effect in 33 patients who took formoterol during pregnancy (Wilton and Shakir, 2002). Moreover, an epidemiologic study reported no adverse outcomes among 65 women who used salmeterol while pregnant (Mann et al., 1996). The use of LABAs as monotherapy was reported in one study that did not find any evidence of an effect on fetal growth in humans (Bracken et al., 2003).

Some studies in animals have shown that theophylline can cause birth defects when given in doses that are many times higher than the human dose (Shibata et al., 2000). Nevertheless, the NAEPP review of two case reports and six clinical studies that included a total of 57,163 pregnant women, of whom 3616 had asthma and 660 had taken theophylline, reassures regarding the safety of theophylline at recommended doses (to serum concentration of 5–12 μg/ml) during pregnancy (National Heart, Lung, and Blood Institute National Asthma Education and Prevention Program Asthma and Pregnancy Working Group, 2005). Unfortunately, however, the ability to clear theophylline from the body may decrease later in pregnancy (Frederiksen et al., 1986); therefore, dosages of theophylline may need to be adjusted as a result of the changing PK as pregnancy progresses, and because theophylline readily crosses the placenta. Use of this medicine in pregnant women may cause unwanted effects, such as fast heartbeat, irritability, jitteriness, or vomiting, in the newborn infant if the amount of medicine in mother's blood were too high (Labovitz and Spector, 1982). In any case, theophylline may induce several side effects in pregnancy such as nausea/gastroesophageal reflux exacerbations, hypertension, delivery inhibition, etc. (Liccardi et al., 2003). No data concerning use of mAChR antagonists in pregnancy are available, although studies in rats, mice, and rabbits showed no birth defects when these medications were used (Liccardi et al., 2003).