Inhaled corticosteroids for asthma: are they all the same?
Raju C. Reddy, MD, Division of Pulmonary and Critical Care Medicine, University of Michigan Medical Center, 109 Zina Pitcher Place, 4062 BSRB Ann Arbor, MI 48109-2200, USA. Tel.: +1 734 615 2871; fax: +1 734 615 2111; e-mail: firstname.lastname@example.org
Objectives: To assess similarities and differences among currently available inhaled corticosteroids (ICS) for treatment of asthma, with special emphasis on factors that may affect the relative safety of these medications.
Methods: PubMed was searched for relevant reviews and original articles. Information from these studies was synthesized and critically assessed.
Results: Differences in corticosteroid formulations and delivery systems can create variations in therapeutic efficacy. Chemical properties of the various corticosteroids may also affect their relative safety. Ciclesonide and beclomethasone dipropionate are administered as prodrugs activated by enzymes present in the lungs but not the oropharynx. Corticosteroid-specific adverse effects in the oropharynx are thus avoided, although formulation-specific effects may remain. Other adverse effects require systemic availability, either via the gastrointestinal tract or the lung. Once they enter the systemic circulation, all ICS are rapidly metabolized by the liver. Oral bioavailability of ICS such as fluticasone, ciclesonide and mometasone is minimal, as a result of their essentially complete first-pass metabolism in the liver. Ciclesonide also undergoes extrahepatic metabolism that eliminates it even more rapidly. Additionally, ciclesonide and mometasone exhibit very high levels of binding to serum proteins that reduces their ability to stimulate glucocorticoid receptors outside the lung.
Conclusions: Despite acting by similar mechanisms, currently available ICS and their delivery systems differ in ways that can potentially affect both safety and therapeutic effectiveness for individual patients.
Inhaled corticosteroid therapy (ICS) revolutionized the management of patients with asthma (1). For the first time, it was possible to provide long-term control of symptoms without the serious systemic side-effects of oral corticosteroids. This was not, however, the first use of inhalers in asthma patients. Metered-dose inhalers had been introduced in the 1950s for delivery of short-acting β-agonist bronchodilators. This was followed approximately a decade later by introduction of dry-powder inhalers for delivery of sodium cromoglycate. The first ICS was beclomethasone, introduced in the early 1970s. Clinical trials quickly demonstrated that patients previously maintained on low-dose oral corticosteroids could be partially or completely transferred to inhaled beclomethasone dipropionate with greater safety and little loss of efficacy (2–4).
Today, ICS are the recommended first-line therapy for persistent asthma of all severities and patients of all ages (5) and are the most effective asthma medications currently available. Limitations and drawbacks remain, however. Whether corticosteroids can inhibit the airway remodelling characteristic of chronic asthma remains controversial. Furthermore, an estimated 5–10% of all asthma patients do not respond adequately to even oral steroids (6–8). Additionally, although side-effects of ICS are less frequent and less severe than those of oral steroids, safety concerns remain.
Potential adverse effects
Effects in the oropharynx
Although use of inhaled as compared with oral steroids dramatically reduces the incidence and severity of systemic adverse effects, they raise the possibility of local adverse effects specifically in the oropharyngeal cavity. A significant fraction of the delivered dose, perhaps as much as 90% (9), may be deposited in the mouth and pharynx without ever reaching the lungs.
The mechanisms underlying local adverse effects of these oropharyngeally deposited corticosteroids have not been extensively investigated, but appear to include both immunologic and non-immunologic actions. It is presumably suppression of local immunity that leads to the increased risk of oral candidiasis seen with use of ICS. The extent of the risk is unclear, however, perhaps in part because Candida colonization is often unassociated with clinical symptoms (10), so studies reporting colonization and those reporting clinical candidiasis are not truly comparable. One study covering almost 26 000 patients age 65 and older who were using ICS found a 3-year oral candidiasis incidence of 7% (11), but whether this frequency would be observed in patients of other ages remains uncertain. Fortunately, the condition is usually easy to treat.
Dysphonia accompanying ICS use is common. The true frequency remains unclear, however, with reported incidences ranging from 5 to 50% (12). The variations in reported incidence may reflect both varying methods of ascertainment, often patient questionnaires or incidental findings, and varying definitions that may or may not differentiate dysphonia from hoarseness of other aetiologies. Steroid-associated dysphonia has sometimes been attributed to myopathy affecting the muscles that control the vocal cords, but laryngoscopic findings have been inconsistent (13). One study found that switching from a metered dose inhaler to a dry powder inhaler significantly reduced the frequency of dysphonia, possibly because of differences in position of the vocal cords during use of the two devices (14).
Persistent cough and bronchospasm are relatively uncommon side-effects of ICS use (13). However, reflex cough during inhalation is quite common. One study observed reflex cough in 34% of a mixed-age population using either beclomethasone dipropionate or budesonide (15). Both this study and another (16) found that use of a large-volume spacer did not reduce the frequency of cough. They likewise found no effect of differences in the medications tested. Lack of any difference among ICS in the frequency of reflex cough could point toward this being due to the propellant or to irritant properties of other components of the formulation.
In principle, steroid effects in the oropharyngeal cavity can be avoided by administering the agent as a prodrug that must be activated with an enzyme present in the lung but not in the upper respiratory tract. Ciclesonide is one such agent and its prodrug status has often been cited as an advantage. A pooled analysis found a lower incidence of oropharyngeal adverse effects compared with fluticasone, budesonide and beclomethasone dipropionate as a group (13). It is difficult to tell from the available data whether differences between ciclesonide and beclomethasone dipropionate, which is also administered as a prodrug, were observed in this analysis. However, as other studies have found little difference between beclomethasone dipropionate and budesonide, the prodrug concept remains open to question (13).
Suppression of the hypothalmic–pituitary–adrenal (HPA) axis
Glucocorticoid secretion is governed by a negative feedback loop in which activation of hypothalamic glucocorticoid receptors results in reduced secretion of corticotrophin-releasing hormone. This leads to a decrease in release of adrenocorticotrophic hormone (ACTH) by the pituitary and consequently of cortisol and other glucocorticoids by the adrenal gland. As exogenous glucocorticoids activate the hypothalamic glucocorticoid receptor, they suppress production of endogenous glucocorticoids. Long-term, high-level suppression can render endogenous production slow to recover, so that sudden withdrawal of the exogenous agent can result in acute adrenal insufficiency and adrenal crisis (17). There has also been concern about possible adverse effects of lesser degrees of HPA axis suppression, although this is controversial.
Even one-time administration of ICS has been shown to produce measurable HPA axis suppression. For example, a 500 μg dose of fluticasone administered via a dry-powder inhaler produced a 29% reduction in nocturnal cortisol secretion compared with the preceding night’s values (18). Another study found that fluticasone doses of 250, 500 and 1000 μg reduced plasma cortisol area under the concentration–time curve (AUC) for the first 20 h after dosing by 8%, 19% and 28%, respectively, while a 800 μg dose of budesonide reduced values by 16% (19). These modest single-dose reductions are unlikely to be clinically significant. However, a cross-sectional study has found clearly abnormal (<5 μg/dL) morning cortisol levels in 10 of 28 patients receiving >660 μg of fluticasone for a year or more (20). All of these patients also exhibited a low response to ACTH stimulation. Furthermore, a survey in the UK found that adrenal crisis had been identified in 33 ICS users among the patients of 709 responding physicians (21). Most of these cases were children receiving fluticasone and clinical practice recommendations have since been changed to eliminate the doses at which adrenal crisis was reported.
Other safety concerns
Skin thinning and ecchymoses. Use of ICS is associated with reduced synthesis of skin collagen (22), which at higher doses leads to skin thinning and ecchymoses (23) as well as to slow healing of skin cuts and sores (24). Frequent bruising has been documented in a number of large studies (24–27), presumably reflecting capillary fragility as a result of impaired collagen synthesis.
Decreased bone mineral density. Effects of ICS on bone mineral density are controversial. Confounding factors include previous use of oral corticosteroids and respiratory-related lifestyle changes, such as limited exercise, that may themselves affect bone mineral density. However, even studies designed to limit or avoid effects of oral corticosteroid use may give conflicting results (28–31), as do studies with biochemical markers of bone formation and degradation (32–34). One particularly well-designed study enrolled only patients aged 20–40 with little or no previous oral corticosteroid use. With major confounding factors controlled, a negative relationship between bone mineral density and cumulative lifetime dose of ICS was observed (35). This is in accord with the consensus of reviewers that long-term, high-dose use of ICS may have significant effects on bone mineral density (36–38), although a Cochrane review disagrees (39). The Cochrane meta-analysis of seven studies meeting inclusion criteria concluded that there was no evidence of increased loss of bone mineral density or of fracture risk and that increases in osteocalcin levels (a measure of bone mineral degradation) were seen only at doses above those usually recommended.
Cataracts and glaucoma. Evidence for an association between ICS use and development of posterior subcapsular cataracts is likewise not completely consistent. One small study found that cataract incidence was associated only with prednisone use (40). Two larger studies, however, have found an association with dose and duration of ICS use (41, 42). A recent meta-analysis has reached the same conclusion, finding that ICS increased cataract risk significantly (43).
One large case–control study found an increased risk of glaucoma or ocular hypertension with high doses of ICS but not with ICS use in general (44). Initiation of ICS use likewise showed no effect on glaucoma incidence in a smaller prospective study (45). Glaucoma thus does not seem to be a general concern in patients receiving ICS, although measurement of intraocular pressure may be appropriate in select patients receiving high doses.
Although safety concerns remain with ICS, these do not offset the established beneficial profile of these drugs. Most strikingly, one study has found that prescribing an ICS to elderly patients being discharged following an asthma-related hospitalization resulted, after adjustment for confounders, in a 29% relative reduction in the risk of rehospitalization and a 39% reduction in the risk of death during the following year (46). Nevertheless, the National Heart, Lung, and Blood Institute’s Expert Panel recommends use of the lowest possible ICS dose that maintains asthma control, with consideration being given to adding a long-acting β-agonist to low- or medium-dose ICS therapy in preference to further escalation of the ICS dose (5). The panel also recommends calcium and vitamin D supplements, and possibly bisphosphonates where indicated.
Pharmacokinetics and pharmacodynamics of ICS
Oral and pulmonary bioavailability
Inhaler-delivered corticosteroids become systemically available by one of two routes. A large fraction is deposited in the oropharyngeal cavity and is then swallowed. Varying amounts will be absorbed into the enterohepatic circulation and subjected to first-pass metabolism in the liver. Oral bioavailability (Table 1), which is defined as the fraction of an oral dose that reaches the systemic circulation, reflects both the fraction absorbed and the fraction escaping first-pass metabolism. As an example, the oral bioavailability of budesonide is approximately 11% (47). The oral bioavailability of beclomethasone, however, is approximately 41%, with 26% of an inhaled dose becoming systemically available via the gastrointestinal tract (48). By contrast, the oral bioavailability of fluticasone and ciclesonide is ≤1%, representing essentially complete first-pass metabolism (49–51).
Table 1. Oral bioavailability of inhaled corticosteroids
| As unchanged compound||<1%|
| As beclomethasone 17-monopropionate|| 41%|
|Budesonide||11–13% (47, 73)|
|Fluticasone||≤1% (50, 78)|
|Mometasone||No data availablea|
Essentially all drug that reaches the lungs and is not removed by mucociliary clearance will eventually enter the systemic circulation. This means that pulmonary bioavailability is determined by inhaler design and the patient’s technique of usage, along with size and physical properties of the particle, but not by the chemical identity of the drug. As typical examples, the pulmonary deposition, and hence bioavailability of ciclesonide using a hydrofluoroalkane inhaler is 52% while that of fluticasone is only 16% using the same inhalation device (52). The difference probably reflects a difference in formulation (solution vs. suspension). Particle size and composition also affect how deeply the particle penetrates into the lung. Distal, even deposition is therapeutically desirable but does not ultimately influence systemic availability.
The absolute bioavailability of any ICS, defined as oral plus pulmonary bioavailability, is always less than 100%. Notably, the absolute bioavailability of fluticasone is only about 16% (50). This reflects a combination of incomplete absorption from the gut and first-pass metabolism.
Pulmonary residence time
As ICS act therapeutically in the lung, it would seem desirable for a drug to reside longer in the lung before entering the systemic circulation. Whether increased pulmonary residence time, defined as the mean time between inhalation and systemic absorption, actually increases the receptor’s total exposure to active drug depends on the mechanisms involved.
One factor affecting pulmonary residence time is the rate at which inhaled particles dissolve. This calls for a careful balance in drug design. A highly soluble drug may dissolve quickly, giving a rapid peak in drug concentration that then falls rapidly as the drug enters the bloodstream. If the drug dissolves too slowly, however, the particles may be swept out of the lung by mucociliary clearance. The rate of dissolution is affected by particle size, the physical characteristics of the inhaled particle and the chemical characteristics of the drug. The possibility that dissolution rate may be modified by such formulations as coated particles and microspheres is currently under investigation.
The other way in which pulmonary residence time can be increased is by esterification of the drug to a fatty acid. These esters are very highly lipophilic and will typically be sequestered in the cell membranes until the ester bond is hydrolysed in the course of normal cell metabolism. However, these membrane-bound esters are not available to activate pulmonary glucocorticoid receptors. Thus, the net effect is to maintain a pulmonary reserve of inactive drug. This ensures relatively steady levels of active drug without increasing total receptor exposure. Formation of these fatty acid esters requires that the drug have an unhindered hydroxyl at position 21. Such a hydroxyl is present in budesonide (53), triamcinolone (54) and the active metabolite of ciclesonide, desisobutyryl-ciclesonide (55). Esters of desisobutyryl-ciclesonide have been shown to persist in the lung for more than 24 h (55).
Beclomethasone dipropionate and ciclesonide are administered as prodrugs that are activated in the lung by esterase activity. Beclomethasone is administered as the 17,21-dipropionate. Since a free hydroxyl group at the 21 position is required for activity, the prodrug is inactive until esterases in the lung convert it to the active 17-monopropionate and later to the less-active parent compound (48). Ciclesonide is a 21-isobutyryl ester that is hydrolysed by the same enzymes to the active desisobutyryl-ciclesonide (56).
The relevant esterases are not present in the oropharynx, so administration as a prodrug is expected to improve oropharyngeal safety. These enzymes are present elsewhere in the body, however, so any drug that escaped activation in the lung will still be activated once it enters the systemic circulation and can potentially contribute to systemic adverse effects. The simultaneous presence of active and inactive forms complicates study of the pharmacokinetics, though.
A highly lipophilic drug passes readily through the cell membrane and thus reaches the glucocorticoid receptor. This is equally true in the lung and in extra-pulmonary tissues. Thus, both benefit and risk are increased without any necessary alteration in their ratio. An additional effect, however, is that a larger fraction of a highly lipophilic drug will leave the bloodstream and be temporarily sequestered in body tissues. In single-dose studies, this will decrease the amount present in the bloodstream. With chronic dosing, however, the drug will accumulate in the tissues and an equilibrium will be reached. Thus, once again, the effect will be to dampen concentration fluctuations without affecting the amount available on average to the receptor.
The apparent volume of distribution is the volume that the total amount of drug in the body would occupy if it were uniformly distributed at the concentration in the blood. It is thus a measure of the extent to which the drug leaves the bloodstream and is sequestered in the tissues. Exchange between blood and tissue, however, is an equilibrium process that reflects the amount of unbound drug in each compartment as well as the ease with which drug passes from one compartment to another. A high level of binding to plasma proteins tends to retain drug in the blood whereas high liopophilicity is among the factors that contribute to tissue entry and to binding and retention in the tissues. Reflecting the balance of these two factors as well as others, measured distribution volumes range from approximately 180 L for budesonide to almost 900 L for the active metabolite of ciclesonide (52). Although apparent volume of distribution influences drug half-life, the important factors determining steady-state safety and efficacy are lipophilicity and protein binding individually rather than their combined effects on volume of distribution.
Once a drug has entered the bloodstream, the rate at which it is removed from the body becomes important. Corticosteroids are metabolized primarily by the liver and the clearance is typically similar to or somewhat lower than the hepatic blood flow rate (52). However, clearance of the active metabolite of ciclesonide considerably exceeds hepatic blood flow, indicating involvement of extrahepatic sites of inactivation (57). As rapid clearance decreases the AUC, such extrahepatic inactivation is desirable and is likely to be a common feature in ICS introduced in the future.
An additional consideration is the extent to which corticosteroids in the bloodstream are bound to plasma proteins, predominantly albumin (Table 2). Protein-bound corticosteroids do not activate the glucocorticoid receptor. One might in principle expect them to also be partially protected from inactivation, but first-pass hepatic clearance of ICS is so efficient that the effect is negligible in practice. Thus, protein binding that renders the drug unavailable during the period between absorption from the lung and the time it reaches the liver is an almost unqualified advantage. Reported protein binding for currently available ICS ranges from 71% for triamcinolone to 99% for ciclesonide and its active metabolite (52). There has been a general trend toward increased protein binding by newer ICS, and it is likely that 97–99% binding will remain standard for any that may be introduced in the future.
Table 2. Protein binding of inhaled corticosteroids (Adapted from (52) except as noted)
Potency (Table 3) is affinity for the receptor, that is, the fraction of receptor bound and activated at a given drug concentration. As the same receptor is responsible for both therapeutic and adverse effects, high potency would not in itself alter the risk/benefit ratio. However, whereas the therapeutic effect of most ICS appears to plateau at doses only modestly above those currently recommended (7) – for example, evidence indicates that there is no therapeutic advantage to increasing the budesonide dosage above 1600 μg/day (58) – this appears to be less true for the newer, more potent corticosteroids fluticasone (59) and ciclesonide (60). The same may be true for mometasone (61), where one study in patients with severe, persistent asthma found that doses ranging up to 1600 μg/day (four times the usual dose) allowed 76% of those patients to completely eliminate usage of oral prednisone (62). Such dose escalation becomes particularly attractive when high potency is combined with low oral bioavailability.
Table 3. Rank order of inhaled corticosteroid potency (highest to lowest)
|Fluticasone (82–84) ≈ Ciclesonide (83)|
|Budesonide (84) ≈ Triamcinolone (84)|
Methods of delivery to the lung
Since the therapeutic target of ICS is the lung, a method for delivery to that site is required. The ideal goal would be to deliver drug specifically to the lung and nowhere else, but unfortunately none of the current methods approach that goal. Nevertheless, there are significant differences among the currently available delivery methods.
There are currently three methods by which corticosteroids can be delivered to the lung: metered dose inhalers (MDIs), dry powder inhalers (DPIs) and nebulizers. MDIs have the drug either dissolved or suspended in a liquefied, pressurized propellant. Pressing the actuator button briefly releases the pressure, converting a measured amount of propellant into an aerosol delivered from the orifice. Older MDIs used chlorofluorocarbons (CFCs) as the propellant, but following adoption of the Montreal Protocol on Substances That Deplete the Ozone Layer those were gradually phased out in favour of the hydrofluoroalkanes (HFAs) used today. MDIs require the patient to inspire as the button is pressed, a technique that many find difficult to master. Indeed, fully three-quarters of patients (63) do not use the correct technique while more than 40% of nurses and faculty at an academic medical center perform at least four of the seven steps incorrectly (64). This is one factor that has led to adoption of large-volume spacers or holding chambers for use with MDIs, since these devices retain the drug within the chamber until the patient breaths in. They also facilitate anatomically correct placement of the device outlet and reduce the shock of cold propellant hitting the back of the throat that sometimes causes patients to halt inspiration. These spacers significantly increase the overall bulk of the device, however, and attention to technique is still required.
With DPIs, it is the patient’s own inspiration that provides the energy for drug delivery. That breath draws air through the drug, formulated as a dry powder. The drug is then entrained in the inspired air, with screens or spinning surfaces used to break up the aggregates that micronized particles would otherwise form. DPIs do not require hand-breath coordination, but do require fairly strong inspiratory force (exactly how strong varies with the specific device). However, even asthma patients with reduced ventilatory volume can usually generate adequate inspiratory force, since the reduced volume primarily affects duration rather than force of inhalation (65).
Nebulizers convert a liquid solution or suspension into an aerosol using either a jet of compressed air or ultrasonic energy. The aerosol plume is then delivered to the patient through either a face mask or a mouthpiece. These devices make minimal demands on patient technique. Even with modern advances, however, they remain relatively bulky and are therefore typically prescribed only for patients unable to use either MDIs or DPIs. A systematic review has found that nebulizers appear at least as effective for ICS delivery as MDIs with large-volume spacers (66).
Delivered particle size is an important consideration in choice of an inhaler. Particles should ideally be between 1 and 5 μm in diameter, because larger particles are likely to be deposited in the oropharynx while very small particles will either be deposited in the upper airways or, if drawn into the lower airways, will be exhaled (67). Since replacement of CFCs by HFAs as the propellant required MDI redesign, many manufacturers utilized the opportunity to optimize design in other respects as well. Thus, while an older CFC-driven MDI deposited only 4–7% of beclomethasone dipropionate leaving the device into the lungs, the newer HFA-driven inhaler deposits 55–60% (9). Older studies comparing DPIs with CFC-driven MDIs are therefore no longer meaningful. The HFA-driven fluticasone inhaler, however, was deliberately designed to have the same particle size distribution and lung deposition as the CFC-driven device in order to maintain dosage comparability. This inhaler already had a relatively fine particle size, however, with the proportion of particles <5 μm in diameter being considerably higher than that seen for the same drug delivered by the Diskhaler DPI (68).
Particle sizes delivered by DPIs also differ between devices. Because of the large number of different designs, however, generalizations are difficult and only a limited number of possible comparisons have been addressed. One review has concluded that the Turbuhaler delivers twice as much of the administered drug to the lung as does the Diskus (69), but the studies cited are complicated by delivery of different drugs from the two devices. Clinical results may be of more interest, especially when the same drug is being administered. One such study found that patients receiving beclomethasone dipropionate via the Diskhaler (similar but not identical to the Diskus) were more likely to have used little or no short-acting β-agonist than were those receiving the same drug via the Rotohaler (70). A comparison of budesonide delivered by Airmax or Turbuhaler found that patients using the Airmax demonstrated a non-signficant trend toward greater improvement in airway hyper-responsiveness but with no difference in forced expiratory volume in 1 s, peak expiratory flow or symptoms (71). Turning to nebulizers, Dahlström et al. (72) found no difference in lung deposition of budesonide as a function of nominal dose for three different jet nebulizers [the Inhalierboy, the LC Jet Plus (Pari Gmbh, Starnberg, Germany) and the MA-2 (Clinova Medical AB, Malmo, Sweden)], although a larger fraction of the dose actually delivered by the Inhalierboy was deposited in the oropharynx.
Although early studies reported that the Turbuhaler delivered twice as much budesonide to the lungs as did a MDI (73), and that this was reflected in patients prescribed a Turbuhaler getting the same or greater benefit with half the dose (65), these studies were all done with the CFC-driven MDI. Current evidence indicates no clinical difference among the various devices (74). Specifically, no differences in effectiveness or safety were seen in relatively short-term randomized controlled trials, although oropharyngeal adverse effects were not examined. However, such studies typically require all enrolled patients to demonstrate an ability to use the device correctly and otherwise gloss over factors that may be important for individual patients. Thus, choosing the correct delivery method for a specific patient calls for knowledge of both the device and the patient. Aside from patient ability and willingness to use a given device, patient preference may also be a factor, as may the availability of a given delivery device for the drug of choice.
Only a limited number of pathways toward safer and more effective ICS are available. Because therapeutic and adverse effects are exerted through the same receptor, greater selectivity through structural modification is not an option. Recently introduced agents exhibit near-maximum first-pass metabolism and protein binding, so further improvement in those respects is likewise not an option. Furthermore, one of those recently introduced drugs, ciclesonide, is a prodrug that remains inactive in the oropharyngeal cavity and that also exhibits relatively prolonged pulmonary residence time.
Nevertheless, approaches to improved ICS are still available. One is so-called ‘soft drugs’ that are rapidly inactivated on leaving the desired site of action. As ciclesonide’s clearance exceeds hepatic blood flow, it could be considered a soft drug. There is no reason, however, to believe that ciclesonide’s rate of extrahepatic metabolism is maximal, so room for improvement presumably exists. Bodor & Buchwald (75) review approaches to development of soft corticosteroids and discuss several that are currently in development for asthma or are approved for other indications. These are predominately esters with inactive hydrolytic metabolites, with prolonged pulmonary residence time being the primary approach to achieve the desired pulmonary selectivity. A more intuitively attractive approach would be to look for drugs metabolized by enzymes present in the bloodstream but not in the lung. This appears to be true of glucocorticoid γ-lactones metabolized by paraoxonase, but reports suggest that such compounds may have proven unsatisfactory.
It has also been suggested that compounds might be developed that selectively activate the glucocorticoid receptor so as to transrepress pro-inflammatory transcription factors such as NF-κB and AP-1 without promoting transcription of genes with glucocorticoid response elements (76). Attempts to develop such compounds have had limited success, however, and it is not even clear that such selective activation is possible.
Although ICS remain the accepted initial therapy for persistent asthma, safety concerns persist. These concerns are especially significant when it becomes desirable to escalate the dose beyond those usually recommended. Cataracts and loss of bone mineral density may be especially significant. Suppression of the HPA axis is often seen but becomes clinically significant, at least in adults, only at extremely high doses. Recent developments have eased some of these concerns, as newer ICS have essentially zero oral bioavailability and exhibit 98–99% binding to serum proteins. Some are also prodrugs that remain inactive in the oropharynx and therefore do not produce local adverse effects. Successful efforts to improve delivery devices are also ongoing. With further research, safer or even virtually risk-free ICS may be anticipated.
This study was supported by National Institutes of Health Grant HL070068 (to R.C.R.).