Purpose of review:Although inhaled corticosteroids (ICSs) are the mainstay of therapy in asthma, their use raises certain safety concerns. We review the articles appearing in the last year which have addressed the safety of ICSs when used in the treatment of asthma and chronic obstructive pulmonary disease (COPD).
Recent findings: Recent studies suggest that patients with asthma as opposed to COPD do not experience an excess risk of pneumonia with ICS use. Patients with respiratory diseases are at increased risk of developing active tuberculosis and this excess risk is exacerbated by the use of high doses of ICSs. ICSs have systemic effects and one result appears to be an increase in the risk of diabetes onset and progression, especially at high doses of ICSs. When examining cases of glaucoma requiring therapy, there was no increase in risk with ICSs even at high current and cumulative doses. Finally, use of even high doses of ICSs during pregnancy does not appear to affect foetal adrenal function.
Summary ICSs are a highly effective therapy in asthma and have an excellent safety profile at the low doses usually required in asthma. Adverse effects appear mostly at higher doses.
Inhaled corticosteroids (ICSs) are the mainstay of asthma therapy at all ages. Their broad anti-inflammatory activity controls the airway inflammation underlying airway hypersensitivity to viral infections, allergens and irritants. ICSs are the only asthma medication that has been shown to reduce mortality and the risk of hospitalization over the long term. Most of the therapeutic benefit is obtained at low doses such as fluticasone 200mg/day or the equivalent, whereas adverse effects increase with increasing dose in an apparently linear fashion. It is therefore important to use the lowest effective dose possible and to attempt to reduce the dose further in patients whose asthma has been well controlled for several months. Such dose lowering is made safer by providing patients with an effective action plan to increase the dose for a short period of time if symptoms worsen.
Although the various ICSs available are all therapeutically effective, certain products present safety advantages worth considering. Budesonide has the most evidence for safety during pregnancy and is therefore the drug of choice for women considering having children. Ciclesonide has substantially less systemic activity than other available ICSs, such that it may be preferred in patients requiring higher doses or in children in whom growth retardation is of significant concern.[5,6•] Furthermore, ciclesonide appears to be associated with fewer upper airway adverse effects such as hoarseness and pharyngeal candidiasis than other ICSs of similar potency such as fluticasone.
ICSs are metabolized via the cytochrome P450 system, which is involved in biotransformation of the vast majority of all drugs currently available.[8,9] Therefore, patients often find themselves taking several medications competing for the same enzymes, which are required for their biotransformation into inactive compounds, which can then be excreted. The cytochrome P450 enzyme most commonly involved in drug metabolism is the CYP3A4 isozyme, which is responsible for the first-pass metabolism by the liver of the commonly used ICSs. Inhibition of this metabolism of ICSs can be temporary and concentration dependent (competitive inhibition) or irreversible through formation of a stable metabolic intermediate through permanent binding of the inhibiting drug with the P450 enzyme. Irreversible inhibitors are of particular relevance because they can decrease the first-pass clearance and functional catalytic activity of drugs that are cleared by CYP3A4 until new enzyme can be manufactured. Examples of commonly used irreversible inhibitors of CYP3A4 are clarithromycin, erythromycin, isoniazid, carbamazepine, tamoxifen, ritonavir, verapamil and fluoxetine. Such interactions may be of particular relevance to patients with chronic obstructive pulmonary disease (COPD), as systemic inflammation appears to be a common accompaniment to COPD and inflammation and infection are associated with a further downregulation of cytochrome P450 enzymes. Severe adrenal insufficiency is now a well recognized adverse outcome of concomitant use of the ICS fluticasone and the potent cytochrome P450 inhibitor ritonavir.[13,14] Macrolide antibiotics (clarithromycin and erythromycin) and the quinolone ciprofloxacin (a strong competitive inhibitor of cytochrome P450 enzymes) are commonly used in patients with respiratory disease. Although less potent inhibitors of cytochrome P450 than ritonavir, the repetitive use of such medications may plausibly be related to an increased risk of adverse effects seen with ICSs.
Inhaled Corticosteroids and Pneumonia
The TORCH (Toward a Revolution in COPD Health) trial, comparing the effect of fluticasone, salmeterol or their combination versus placebo on mortality in COPD patients, was the first to report an excess of pneumonia in patients receiving the ICS fluticasone. In a large observational study of over 175 000 patients identified using health administrative databases from the province of Quebec, we confirmed this excess. Patients with COPD who had a current prescription for an ICS were 70% more likely to be hospitalized with a primary diagnosis of pneumonia: relative risk (RR) 1.70, 95% confidence interval (CI) 1.63–1.77. The risk was greatest for those receiving the highest doses equivalent to fluticasone 1000 mg/day, RR 2.25, 95% CI 2.07–2.44. An excess of pneumonia was also confirmed in a meta-analysis of clinical trials of ICSs in COPD. Surprisingly, a meta-analysis of individual patient data restricted to clinical trials of budesonide in patients with COPD did not find an excess of pneumonia. Whether or not there is an excess in mortality resulting from pneumonia contracted by patients with COPD who are prescribed ICSs appears to be controversial. In the large observational study we carried out in Quebec, we found that ICS was associated with a 53% (95% CI 30–80%) excess in pneumonia hospitalization ending in death within 30 days. Inhospital mortality was not different between those prescribed and not prescribed ICS before the event. No excess in total mortality was found in the metaanalysis of clinical trials of ICSs in COPD. This may reflect the younger, healthier clinical trial populations and the lack of information on causespecific mortality. In a reanalysis of the TORCH study, a greater number of pneumonia deaths was seen in the group assigned to fluticasone alone, although no firm conclusion could be made because of the small number of events. A recent report from the Veterans Administration in the USA found that among patients with a listed diagnosis of COPD hospitalized for pneumonia, those who had been prescribed ICSs as outpatients actually had a lower in-hospital mortality. This is ikely due to differences in type and severity of respiratory disease in patients prescribed ICSs. Specifically, such patients are more likely to have an asthmatic component to their obstructive lung disease and this is associated with a better prognosis. Recently, Singanayagam et al. [22•] carried out a careful study of community-acquired pneumonia in patients with COPD confirmed by spirometry and found no difference in mortality after a hospitalization in relation to prior ICS use, whether alone or combined with a beta-agonist bronchodilator. Thus, although in-hospital mortality is likely unrelated to prior treatment with ICSs, the risk of severe pneumonia is increased and therefore, at least in our study, there is a clear excess of pneumonia mortality with ICSs.
It is easy to conceive that ICSs might be associated with an excess in pneumonia, given the chronic bacterial colonization found in many patients with COPD. Furthermore, there is evidence that corticosteroids reduce local defences in peripheral airways of patients with COPD. One would expect that patients with asthma might behave differently. O'Byrne et al. [24•] carried out a metaanalysis of clinical trials of budesonide compared to placebo or to fluticasone among patients with asthma. There was no excess of pneumonia reported as a severe adverse event with budesonide and there was no difference in the risk of pneumonia between budesonide and fluticasone in patients with asthma. When including milder events, budesonide was actually protective. This protective effect is likely explained by a reduction in asthma exacerbations associated with mucus plugging and atelectasis that might be reported as pneumonia radiographically.
Inhaled Corticosteroids and Tuberculosis
Oral corticosteroids are a recognized risk factor for the development of active tuberculosis. Our group investigated whether the risk of tuberculosis was also increased in users of ICSs using the Quebec healthcare administrative databases.[25•] We identified a cohort of 427 648 patients treated with respiratory medications who were on average 52 years of age (±27 years) at cohort entry. Overall, the risk of active tuberculosis amongst this population of patients with respiratory disease was nearly four times that of the general population. Current users of ICSs were slightly more likely to develop tuberculosis, RR 1.33, 95% CI 1.04–1.71. This risk was confined to patients who had not received oral corticosteroids in the prior year, such that ICSs did not confer an additional risk to that of oral corticosteroids. Amongst patients who had not been dispensed oral corticosteroids in the last year, the risk of active tuberculosis was increased almost twofold with higher doses of ICSs equivalent to 1000μg or more per day of fluticasone (RR 1.97, 95% CI 1.18–3.30).
Inhaled Corticosteroids and Diabetes
Studies from the 1990s did not find an association between the use of ICSs and risk of diabetes. Lower doses of ICSs were being used at that time and these medications were used less extensively in elderly patients with COPD who are at greater risk of diabetes. We therefore re-examined the risk of diabetes onset and progression in a large cohort of patients prescribed respiratory medications followed up to the end of 2007.[27•] Amongst 388 584 patients without prior treatment for diabetes, there was a 34% increase in the risk of newonset diabetes (RR 1.34, 95% CI 1.29–1.39). Risk was greatest at higher doses of ICSs equivalent to 1000 μg or more per day of fluticasone (RR 1.64, 95% CI 1.52–1.76). Amongst 30 167 patients who developed diabetes during follow-up, 2099 progressed from oral agents to insulin. The risk of such a progression was also higher amongst patients dispensed ICSs (RR 1.43, 95% CI 1.17– 1.53).
Inhaled Corticosteroids and Glaucoma
Glaucoma is the leading cause of blindness worldwide. Raised intraocular pressure is a risk factor for glaucoma and ocular and systemic corticosteroids increase intraocular pressure. Whether or not ICSs raise intraocular pressure is less clear. An earlier study by Garbe et al.  found an increased risk of glaucoma among users of ICSs. Nearly half of the glaucoma cases were not associated with any therapy, however, thus putting into question the validity and importance of these diagnoses. Furthermore, this earlier study was carried out before the availability of the newer more potent ICSs such as fluticasone. We therefore carried out a large observation cohort study based on the health administrative databases of the Province of Quebec [29•]. Cases were chosen from amongst 196 964 patients 66 years of age or older dispensed respiratory medications. Cases were patients who had visited an ophthalmologist in the prior 90 days and who had been dispensed medication for glaucoma or underwent surgery for this condition. Controls were patients who had visited an ophthalmologist within the same time period but who did not receive a glaucoma diagnosis or treatment for this condition. Therewas no difference in the use of ICSs between the 2991 cases and the 13 445 controls (RR1.05,95%CI 0.91–1.20).Noexcess riskcouldbe demonstrated at the highest doses of ICSs equivalent to 1000 μg or more per day of fluticasone, nor when limiting the analysis to patients who had or had not been dispensed oral corticosteroids. We also examined the effect of cumulative dose of ICSs and found no increase in risk of glaucoma. Therefore, ICSs may be used without incurring an increase in risk of glaucoma requiring therapy. This is in contrast to the increased risk of cataractswhich is seen even with low doses of ICSs in the elderly.[30,31]
Inhaled Corticosteroids and Pregnancy
Poorly controlled asthma is associated with poor pregnancy outcomes including preeclampsia, preterm birth and low birth weight. These can be prevented with regular use of ICSs, and therefore ICSs during pregnancy are recommended in current guidelines. Despite this, there is lingering concern regarding the safety of ICSs during pregnancy. As an illustration, a recent report linked the use of high doses of ICSs dispensed in the first trimester of pregnancy to an increase in the risk of congenital malformations. An earlier report from the same group had not found an excess risk at lower doses of ICSs. A report published this year examined the effect of ICSs use during pregnancy and maternal and foetal hormone levels.[36•] Although a dosedependent suppression of maternal hormone levels was seen (cortisol, osteocalcin), there was no effect on foetal adrenal function even at high maternal doses of ICSs.
ICSs are a highly effective therapy in asthma and have an excellent safety profile at the low doses usually required in asthma. Adverse effects appear mostly at higher doses which can usually be avoided by using combinations of different therapeutic classes and decreasing the dose of ICSs once asthma control has been achieved. If high doses of ICSs are required for prolonged periods, consideration should be given to using an ICS with less systemic activity such as ciclesonide.
- In contrast to patients with chronic obstructive pulmonary disease, patients with asthma prescribed inhaled corticosteroids (ICSs) do not appear to be at increased risk of pneumonia.
- High doses of ICSs increase the risk of reactivation of tuberculosis, although this is not additive to the risk of oral corticosteroids.
- High doses of ICSs increase the risk of clinically overt diabetes and the progression to insulin therapy.
- ICSs, even at high doses, do not seem to increase the risk of glaucoma requiring therapy.
- Papers of particular interest, published within the annual period of review, have been highlighted as:
•of special interest
••of outstanding interest Additional references related to this topic can also be found in the Current World Literature section in this issue (pp.93-94).