Inhaled ciclesonide versus inhaled budesonide or inhaled beclomethasone or inhaled fluticasone for chronic asthma in adults : a systematic review

CRD summary This review evaluated ciclesonide and alternative inhaled corticosteroids for asthma in adults. It concluded that there is limited evidence that inhaled ciclesonide has similar effectiveness and efficacy to fluticasone and budesonide, but no clear evidence on side-effects. The conclusions about limited evidence seem reliable; further data may become available when identified studies reported as abstracts are published in full.


Background
Inhaled corticosteroids (ICS) have a central role in the treatment of asthma. They are the most effective prophylactic agents available, particularly in patients with mild to moderate asthma and persistent symptoms [1] and are recommended for most adult patients with chronic asthma whose symptoms are not controlled by inhaled short acting β 2 agonists [2][3][4]. Regular treatment with corticosteroids reduces exacerbations, improves control of symptoms and lung function, while reducing hospital admissions and deaths from asthma [1,5]. However, prolonged use in persistent asthma and increased doses in severe cases may result in suppression of the hypothalamic pituitary adrenal (HPA) system with concern that this might cause growth impairment in children (including premature closure of the epiphyses of long bones), disturbed glucose tolerance, decreased mineralisation of bone (increasing the risk of fractures), ocular problems such as glaucoma and cataracts as well as thinning of the skin [6][7][8]. Local adverse affects, even at low doses may include dysphonia, pharyngitis and oral candidiasis [2,9].
The mainstay of ICS treatment has been with three agents: beclomethasone dipropionate (BDP), budesonide (BUD) and, more recently, fluticasone propionate (FP). All these agents are similar chemically and structurally but have different pharmacodynamic properties resulting in different clinical effects [10]. A Cochrane review of 48 studies comparing the three agents concluded that FP given at half the daily dose of BDP or BUD leads to small improvement in measures of airway calibre (peak expiratory flow (PEF) and forced expiratory volume in one second (FEV1) while at the same daily dose FP appears to have a higher risk of side effects than BDP or BUD [10].
Ciclesonide (CIC) is a new ICS, manufactured by Altana Pharma Ltd. CIC is licensed only for the treatment of persistent asthma in adults (18 years older) and is delivered via a hydrofluoroalkane metered-dose inhaler (HFA MDI) in 40, 80 and 160 mcg formulations. The recommended starting dose is 160 mcg given in the evening, with reduction to 80 mcg for maintenance [11]. These doses are "exactuator", i.e. the dose expelled, as opposed to "ex-valve" i.e. the actual dose contained in the inhaler (for consistency, throughout this review ex-valve doses are used for all inhalers).
Ciclesonide has little anti-inflammatory activity itself and requires cleavage by endogenous carboxyl esterases in the lung, which creates the active metabolite desisobutrylciclesonide (des-CIC) [12]. This targets activity at the desired location. Des-Ciclesonide undergoes rapid hepatic metabolism into inactive metabolites on leaving the lung [13]. These factors, together with the fact that ciclesonide has very low oral bioavailability due to almost complete first pass metabolism [14] would seem to create conditions favouring the maximisation of therapeutic effect in the lung and minimisation of the risk of systemic adverse effects.
Given that ciclesonide is being actively marketed as an alternative to alternative to other inhaled corticosteroids, our objective in this study was to systematically review published randomised controlled trials of the effectiveness and safety of ciclesonide compared to alternative inhaled corticosteroids in people with asthma.

Search strategy
We performed literature searches on MEDLINE (from 1951), EMBASE, PUBMED and the COCHRANE LIBRARY (The Cochrane Central Register of Controlled Trials) using the following terms: Ciclesonide, Alvesco, Budesonide, Pulmicort, Beclomethasone, Becotide, Becloforte, Fluticasone, Flixotide We imported abstracts of citations from this search into an electronic database. We also searched for "ciclesonide" and "alvesco" on the web sites of the following internet evidence sources: Reference lists of retrieved articles were scrutinized for further studies but yielded no additional papers.
Titles and abstracts were sifted by two reviewers and prior to retrieval of full articles, the two reviewers independently assessed papers for inclusion. Cases of disagreement were resolved by discussion (or, where necessary, delayed until retrieval of the full text). All full text papers retrieved were again assessed by the two independent reviewers for inclusion and methodological quality (see Appendix 1). There was no blinding to authors' names or institutions and no scoring system (such as the Jadad score) was used. Data were extracted by one reviewer.

Inclusion criteria for considering studies Participants
Studies in human adults (i.e. people aged 18 and over) with a diagnosis of chronic asthma were included. We did not consider studies concerning acute asthma, chronic obstructive pulmonary disease or allergic rhinitis.

Intervention
Studies concerning only inhalation of ciclesonide and excluded those involving oral, nasal or intravenous routes.

Comparators
Studies that compared ciclesonide to either budesonide or beclomethasone or fluticasone and excluded studies comparing to placebo only or to other asthma treatments.
Outcome measures we considered all reported outcomes although prominence was given to patient based outcomes.

Study design
Prospective, randomized, controlled trials and or reviews that were clearly systematic and carried out since the beginning of 2004. Studies published only as abstracts were included only if they contained sufficient methodological detail to enable critical appraisal. We considered studies in all languages. Table 1 shows the results of the search. A total of five RCTs were included (details in Table 2) [15][16][17][18][19]. No reviews or abstracts qualified for inclusion. Two RCTs compared CIC to BUD [15,19] and three RCTs compared CIC to FP [16][17][18]. The Kanniess et al study [19] was the only RCT not apparently sponsored by a pharmaceutical company manufacturing CIC.

Quality of the evidence
All papers were critically appraised for methodological quality based on the criteria shown in Table 3. All trials involved small numbers (the largest number of participants completing the study being 19 [18]) and were of very short duration (maximum four weeks). Three of the studies had drop-outs after randomization with attrition rates varying from 5.25% -30% [17][18][19]. In general, pop-ulations were similar, although two studies did not exclude smokers. Participants had mild asthma with mean FEV1 greater than 90% of the predicted value in three of the trials.
Three of the studies were cross-over trials and washout periods were all of appropriate length [17][18][19]. Nave et al and Richter et al were" within patient" trials, and therefore similar to a crossover design, but both treatments were given concurrently, presumably on the assumption that oropharyngeal deposition was independent of treatment but potentially confounded by short term changes in oropharyngeal conditions. There were no parallel studies. There was insufficient reporting to verify whether there was good balance at baseline in the crossover trials. In the Kanniess et al study there was also imbalance between the population as a whole before receiving CIC and the same population before receiving BUD.
Only one trial (Kanniess et al) was clearly double (or more) blinded. Lee, Fardon et al and Lee, Haggart et al it state that the inhalers were "masked" but it is not clear whether their identity was withheld from the treatment administrator or observer as well as the patient. None of the trials reported whether, or how, they concealed allocation of treatment within participants or methods for randomization. Overall there was no evidence of performance bias.
Only two trials (Lee, Fardon et al and Lee, Haggert et al) measured patient based outcomes as end points i.e. symptoms and QoL. All the others measured intermediate outcomes and any interpretation of these results will require an assessment of the degree to which these outcomes are clinically significant.
Some studies (e.g. Kanniess et al [19]and to some extent Lee, Fardon et al [17]) only reported pre and post treatment results within the same treatment (i.e. CIC or the comparator). Although the authors stated there was no significant difference between treatments, they showed no data or calculations.  [20] Airway responsiveness to provoking agent • This is measured in terms of the concentration of inhaled provoking agent (adenosine monophosphate (AMP) or metacholine) required to cause a 20% fall in FEV1 (PC20). The initial dose is inhaled and the FEV1 measured subsequently. The dose is then doubled progressively until a 20% fall is recorded • Plasma cortisol response to human corticotrophinreleasing factor (hCRF). This test has been shown to detect   impaired adrenal reserves in corticosteroid-treated patients [21].

• Urine cortisol
Oropharyngeal deposition • This is the amount of inhaled steroid that does not enter the lung and is deposited in the oropharynx. Table 4 shows the comparative results from the trials.

Results from the trials
None of the trials showed CIC to have any benefit over either FP or BUD for the outcomes of lung function, symptoms, quality of life, airway responsiveness to a provoking agent or inflammatory markers.
Lee, Fardon et al studied HPA suppression. At the end of each four week treatment period, of either CIC 800 mcg bd or FP 1000 mcg bd, a 10 hour overnight urine collection (OUC) was taken and plasma cortisol response to a 100 mcg bolus of hCRF was assessed at 30 and 60 minutes. The authors state that data were logarithmically transformed to normalize the distribution but give no comment on how the data were skewed. Results comparing CIC and FP are reported as the geometric mean fold difference (GMFD) but there is no explanation as to how these values were calculated. By definition a GMFD of 1.0 means no difference.
When comparing the two treatments there was no significant difference in outcome with respect to plasma cortisol response to hCRF. However, there was significantly more suppression of urinary cortisol after FP than CIC (but with 95% CI of 1.1-2.0 this was only barely so). The results for plasma cortisol levels 60 minutes after hCRF are not reported in the table but the authors state that there was no significant difference between pretreatment and post treatment FP levels.
Richter et al compared CIC 800 mcg to FP 1000 mcg each via HFA MDI. Curves were plotted for recovery of each drug in rinsing solution against time after administration and then the molar area under the curve for 0 -60 minutes (AUC 0-60 min ) was calculated for CIC, des-CIC and BUD (or FP) to allow direct comparisons.
The Nave et al study shows that the combined deposition of CIC and des-CIC in the oropharynx was less than half (47%) of that of BUD. Only 8% of the CIC deposited was converted into the active metabolite des-CIC (suggesting a lack of converting esterases in the oropharynx). Overall the concentration of des-CIC in the oropharynx 60 minutes after inhalation was only 4% of the BUD concentration (i.e. 25 times more BUD than des-CIC).
The Richter et al study shows that the combined deposition of CIC and des-CIC in the oropharynx was only 53% of that of FP. Furthermore only 17% of the CIC deposited was converted into the active metabolite des-CIC. The concentration of des-CIC in the oropharynx 60 minutes after inhalation was only 8% of the FP concentration (i.e. 12.5 times more FP than des-CIC).

Discussion
There are few data directly comparing CIC to other ICS and no published evidence directly comparing CIC to BDP specifically. None of the RCTs showed CIC to offer any benefit over BUD or FP for effectiveness i.e. none of the RCTs showed CIC to offer any benefit over BUD or FP for any patient based outcomes (asthma symptoms or QoL in these trials). Furthermore none of the trials demonstrated any benefit from CIC over BUD or FP for indirect outcomes of efficacy i.e. lung function, improving response to AMP or metacholine as provoking agents or for decreasing markers of inflammation.
All but one of the trials were sponsored by drug companies manufacturing CIC and seem to endeavour to demonstrate CIC to have equivalent efficacy to other ICS but with an improved safety profile. However, none of the studies report analyses which exclude superiority of one treatment over another (hence it is not possible to conclude that CIC was equivalent to FP or BUD for any efficacy outcomes) and the evidence regarding safety is not conclusive.
The conflicting evidence from the Lee, Fardon et al trial might indicate that CIC has less systemic adverse effects than FP. Challenges to this conclusion, however, are twofold. The first comes from the trial itself. This is the only published trial comparing HPA suppression between CIC and other ICS and the results were not unequivocal. There were also some methodological weaknesses in the trial. There was no evidence of concealment of allocation, an attrition rate of 30%, no evidence of blinding other than the participants and the choice of a comparator (i.e. FP) that is reported to have the highest risk of side effects [10] (there is no published evidence directly comparing HPA suppression after treatment with CIC to either BUD or BDP).
The second challenge relates to the correlation between the intermediate outcome of HPA suppression measured at four weeks after the start of treatment with ICS correlates and clinically adverse effects for patients. HPA suppression is a well reported outcome of both short and long term ICS use [22][23][24][25]. However the clinical significance of such suppression is uncertain [6,22,25]. Current evidence suggests that ICS do not cause important systemic side effects in doses of up to 400 mcg/day in children and 800 mcg/day in adults, [26] and even in doses of more than 1 mg/day there is no conclusive evidence that patients are at any increased risk from side-effects [25,23]. Hanania et al report HPA suppression and decreased bone density after regular use of conventional doses of ICS for asthma [24] but there is no conclusive evidence of a clinically adverse effect e.g. bone fractures.
Further long term studies are required to determine the long term risk of clinically significant adverse effects as a result of HPA suppression associated with ICS use in general and specifically with CIC. In the meantime it is not possible to conclude that CIC offers any benefit over other ICS in terms of systemic adverse effects.
With respect to local adverse effects there are similar challenges. Although, CIC might be expected to have fewer local adverse effects (due to the inhaled agent being its inactive metabolite des-CIC) there is no logical explanation why CIC should be deposited in the oropharynx in such smaller amounts than FP or BUD. In the Nave et al study the aut point out that the difference in deposition could be due to the different inhaler de used. HFA MDIs have been shown to produce ICS with a smaller particle size than MDIs [27] resulting in 17% of a 200 mcg dose of BUD being respirable [15] compare 48% of a 200 mcg dose of CIC [8,13]. However, in the Richer et al study both inhalers HFA MDI and the authors make no mention of why deposition might be less given both treatments are inhaled via the same device. Neither trial was blinded in any which could have been a source of bias, and the lack of a logical explanation for vastly different deposition rates makes it difficult to draw any definite conclusion. Fu studies (preferably parallel) with larger populations are required before conclu whether CIC offers any benefit in terms of local adverse effects In addition to the RCTs outlined above there are number of abstracts that have no been published as full papers. These trials involved substantially larger numbers of ticipants and ran for longer duration but have not been included in the analysis since contained insufficient detail for critical appraisal of methodological quality. The re reported in these abstracts do not alter the conclusions drawn from the full papers bu reported for interest in Appendix 1.  Irrespective of any clinical benefit or not CIC is more expensive compared to BEC, BUD and FP as shown in Table 5. Treatment with CIC would come at substantial financial cost since at high dose (1000 mcg daily) CIC is 5.13 as expensive as BDP, 2.27 times as expensive as BUD (800 mcg daily) and 1.39 times as expensive as FP.
Any advantage that CIC might have over existing, cheaper, ICS is predicated on assertions regarding the long term dangers of ICS use. "Steroid phobia" is recognised in other fields [28] and is likely to form the basis for effective direct to patient marketing of CIC, where such advertising is permitted. However, the evidence base on long term inhaled steroid use is far from certain and it is not clear whether the dangers are such that the precautionary principle is justified.
Although it is clear that the evidence base for ciclesonide will expand considerably with the publication of the larger studies excluded from this review, we believe it is important to highlight the limited nature of the evidence base that is currently available for scrutiny by clinicians and policy makers seeking to practice and support evidence based medicine.

Conclusion
There is very little evidence that has been published in full comparing CIC to other ICS. Current evidence is restricted to very small, phase II studies of low power. These demonstrate CIC has similar effectiveness and efficacy to FP and BUD (though equivalence is not certain) and findings regarding oral deposition and HPA suppression are inconclusive. There is no direct comparative evidence that CIC causes fewer side effects since none of the studies reported patient-based outcomes. Treatment with CIC would also come at substantial financial cost compared to other ICS. • Hansel et al report no significant changes from baseline for urine cortisol levels after CIC but a significant decrease after BUD. However they do not compare the two treatments.
• Derom et al report no significant changes from baseline for urine cortisol levels after CIC but a significant decrease after FP. However they do not compare the two treatments.

Summary
None of the results reported in the abstracts challenge the conclusions of the review.