Volume 115, Issue 5 , Pages 940-945, May 2005
Short-term lower-leg growth rate and urine cortisol excretion in children treated with ciclesonide
Article Outline
Background
Measurement of short-term lower-leg growth rate in children by means of knemometry has become established as an integral part of the available measures of systemic activity of topical steroids in children.
Objective
We sought to determine the effects of clinically effective doses of the novel inhaled corticosteroid ciclesonide on lower-leg growth rate and hypothalamic-pituitary-adrenal axis function in children with asthma.
Methods
In a double-blind, placebo-controlled, 4-period crossover study, 24 children aged 6 to 12 years sequentially received ciclesonide (40, 80, and 160 μg) in randomized order once daily in the evening. Each 2-week treatment period was followed by a 2-week washout period. Knemometry was performed at the beginning and end of each treatment period. Cortisol levels in 12-hour overnight urine were measured at the end of each treatment period.
Results
No statistically significant differences were seen in lower-leg growth rates between any of the ciclesonide treatments and placebo. Lower-leg growth rates were 0.412 mm/wk for placebo, 0.425 mm/wk for 40 μg of ciclesonide, 0.397 mm/wk for 80 μg of ciclesonide, and 0.370 mm/wk for 160 μg of ciclesonide. There was no statistically significant dose-response effect. Likewise, no statistically significant differences or dose-response effects were found for urinary cortisol adjusted for creatinine.
Conclusion
Short-term lower-leg growth rate and hypothalamic-pituitary-adrenal axis function are not affected by treatment with ciclesonide in doses intended for clinical use in children.
Key words: Asthma, ciclesonide, knemometry, lower-leg growth rate
Abbreviations used: ANCOVA, Analysis of covariance, Des-CIC, Desisobutyryl-ciclesonide, HFA, Hydroflouroalkane, ICS, Inhaled corticosteroid
Inhaled corticosteroids (ICSs) are becoming increasingly useful for prophylaxis of childhood asthma. Therefore new ICSs with high topical potency and lower systemic effects are being developed. Ciclesonide is the latest example of this development. Ciclesonide is an ester parent compound that has almost no affinity for the glucocorticoid receptor itself but is converted by esterases in the airways and the lungs to the potent and pharmacologically active metabolite desisobutyryl-ciclesonide (des-CIC). A number of in vitro and in vivo studies have shown des-CIC to possess high local anti-inflammatory effects in the airways.1, 2, 3 Studies in adults have found a very low incidence of systemic effects and low concentrations of freely circulating ciclesonide and des-CIC, even after inhalation of high doses of ciclesonide.1, 4, 5 These findings might be explained by a high systemic clearance and protein binding in the blood.6 Moreover, because of the low receptor affinity of the parent compound ciclesonide, local side effects in the oropharynx, such as candidiasis and hoarseness, are rare.5 Ciclesonide has been shown to be clinically very effective at low doses in both children and adults.7, 8, 9, 10, 11 However, few complete assessments of its systemic effects in children are available.
Recently, knemometry has become established as an integral part of the available measures of systemic activity of topical steroids in children.12 By measuring changes in lower-leg length with an accuracy of 0.1 mm, knemometry provides a powerful tool for investigating the influence of exogenous glucocorticosteroids on short-term linear lower-leg growth. In addition, measurement of cortisol excretion in the urine has also been shown to be a sensitive marker of the systemic effects of exogenous steroids.
The aim of this study was to assess the systemic effects of different doses of inhaled ciclesonide by measuring short-term linear lower-leg growth rates by means of knemometry and 12-hour overnight urine cortisol excretion in children with mild asthma during treatment with ciclesonide administered through a hydrofluoroalkane (HFA) metered-dose inhaler.
Methods
Patients
Outpatient children 6 to 12 years of age without any signs of puberty but with mild asthma for at least 6 months (FEV1 ≥80% of predicted value) who required only rescue medication with inhaled β2-agonists for the 3 weeks before the trial, who had no asthma exacerbation or relevant respiratory tract infection for at least 1 month before the study, and who did not have any other concomitant disease apart from asthma and other atopic diseases were enrolled in the study. The study was conducted in agreement with the guidelines for good clinical practice and was approved by the local ethics committee. Informed consent was obtained from all children and their parents before any study-related procedures were undertaken.
Study design and treatments
This was a randomized, double-blind, crossover trial with 4 treatment periods and 3 washout periods (Fig 1). After a 1-week run-in period, during which the children were familiarized with the study procedures and inhaler use, each child was randomized to receive placebo or 40, 80, or 160 μg/d ciclesonide (each child took 1 inhalation of ciclesonide or placebo from an HFA metered-dose inhaler each evening). Each treatment period was 2 weeks long and separated from the other periods by a 2-week washout period. Treatment order and dose were allocated by using a computerized randomization scheme prepared in balanced blocks. Throughout the entire study, children used inhaled β2-agonists as needed. No other medication was allowed. Before the study, all children were taught the correct inhalation technique with HFA inhalers, and the inhalation technique was evaluated at each visit. The patients and their parents were asked to fill out an adherence form in which deviations from the dosage stipulated in the protocol (eg, missed doses) and the respective dates were entered.

Figure 1.
Study design. Randomization schedules with ciclesonide treatment regimens and washout durations for the study period.
Knemometry
Knemometry of the right lower leg was scheduled at the beginning and end of each period by using a knemometer manufactured by the inventor.13 The same trained observer who was blinded to the recordings of the previous visit performed all measurements. The children were measured at roughly the same time (within 30 minutes) in the afternoon (between 1 and 7 pm), as recommended for knemometry. The patients were prohibited from exercising for 2 hours before the visit. At each visit, 4 estimates of the lower-leg length were made. The most deviant value was disregarded, and the mean of the remaining measurements was used for analysis. The technical error of the knemometer (the mean SD of 3 successive estimations of lower-leg length) was 0.07 mm. In addition, height (Harpenden stadiometer; Harpenden Ltd, Crymych, United Kingdom) and weight (electronic beam analyzer) were recorded.
Measures of pulmonary function
At each visit, pulmonary function, namely, FEV1, was evaluated according to American Thoracic Society recommendations. Throughout the study, peak expiratory flow rate was measured at home in the morning and evening (best of 3 efforts with an Astech peak flow meter). In addition, the use of rescue medication (inhaled β2-agonists) and asthma symptoms during the night and day were recorded in diaries. The asthma score was based on a 5-point scale, in which a score of 0 represented no asthma-related symptoms and a score of 4 represented the highest discomfort resulting from asthma-related symptoms (eg, awake most of the night because of asthma or unable to carry out daytime activities because of asthma).
Cortisol measurement
At the end of each treatment period, overnight 12-hour urine samples were collected for analysis of urine cortisol, urine 6-β-hydroxycortisol, and creatinine excretion. Cortisol in the urine was measured by using an HPLC method with an intra-assay and interassay coefficient of variation of 1.8% or less and 4.6% or less, respectively.
Safety
A complete physical examination was carried out at enrollment and at the end of the final treatment period to detect any illness or disorder that might interfere with linear growth. In all patients reporting oropharyngeal events, an oropharyngeal examination was performed. If a fungal infection of the mouth or throat was suspected, a culture was taken and sent to the laboratory for confirmation of the diagnosis.
Statistical analysis
Growth velocities of the lower leg were calculated for each period as the difference between the value at the end of each period (end point) and the corresponding value at the start of the period and expressed as millimeters per week. An analysis of covariance (ANCOVA) model, including treatment, sex, period, and patient within sequence as factors and age and the baseline value of the length of the lower leg as covariates, was used for the 4-period crossover design. Least squares means and 95% confidence limits were calculated for the differences in growth velocity within each treatment group. The differences between treatment groups were compared by using the 95% CIs for effects of 40, 80, and 160 μg/d ciclesonide versus placebo. A trend test on the basis of the regression model described above also was performed to assess dose dependency of the primary variable. The overall level of significance was set to 5% (2-sided). The type III sum of squares was used for inference. Differences between treatment groups with regard to height and FEV1 were assessed by using the ANCOVA model. The effect on morning and evening peak expiratory flow between the treatment groups and the urine cortisol variables were assessed by using the ANCOVA model without the covariate baseline value. The pairwise effect on symptom scores and rescue medication use between the treatments groups was assessed with the Wilcoxon Pratt test.
Results
Patients
All 24 enrolled patients completed the study. Table I summarizes the patient demographics and baseline disease characteristics. The mean FEV1 percent predicted was 92.4% ± 12.2% (from the start of the treatment period), with a mean asthma duration of 8.6 years. Concomitant disease in the study group consisted of allergic rhinitis (58%), allergic conjunctivitis (54%), and atopic dermatitis (17%). Reported adherence to therapy was high (median, 100%; range, 96% to 103%) and was not statistically significantly different between the various treatment periods.
Table I. Patient demographics and baseline disease characteristics
| Variable | Full analysis set |
|---|---|
| Patients, n | 24 |
| Sex, n | |
| 8 | |
| 16 | |
| Median age, y (range) | 10.0 (7-12) |
| Mean height, cm (SD) | 138.9 (9.5) |
| Mean weight, kg (SD) | 34.3 (9.0) |
| Mean FEV1, percent predicted (SD) | 92.4 (12.2) |
| Asthma duration, y (SD) | 8.6 (2.6) |
| Patients with concomitant diseases, n (%)∗ | |
| 14 (58) | |
| 13 (54) | |
| 4 (17) |
∗Percentage of patients based on the total number of patients in the full analysis set. |
Knemometry results
No statistically significant differences were seen in lower-leg growth rates between any of the ciclesonide treatments and placebo. Mean lower-leg growth rate for placebo was 0.41 mm/wk (Fig 2 and Table II). For 40 μg/d ciclesonide, the lower-leg growth rate was 0.43 mm/wk; for 80 μg/d ciclesonide, the lower-leg growth was 0.40 mm/wk; and for 160 μg/d ciclesonide, the lower-leg growth rate was 0.37 mm/wk. No significant dose-response effects were observed (P
=
.42) in lower-leg growth rates between ciclesonide treatments and placebo.

Figure 2.
Effect of ciclesonide on lower-leg growth rate. Individual growth values are shown for each treatment regimen. Horizontal lines represent mean growth values during each treatment.
Table II. Lower-leg growth velocity (in millimeters per week): Differences between treatment groups
| Test - reference∗ | ||||
|---|---|---|---|---|
| Test | Reference | LS Mean ± SE† | 95% CI | P value |
| Ciclesonide, 40 μg/d | Placebo | 0.013 ± 0.060 | −0.105 to 0.132 | .82 |
| Ciclesonide, 80 μg/d | Placebo | −0.014 ± 0.060 | −0.133 to 0.104 | .81 |
| Ciclesonide, 160 μg/d | Placebo | −0.041 ± 0.060 | −0.160 to 0.077 | .49 |
∗Based on ANCOVA model with factors of treatment, sex, period, and patient (sequence) and covariates of age and baseline value. |
†Mean of the difference between the treatment and placebo value. |
Height measurements
Mean change from baseline in height was approximately 0.1 cm/wk for each treatment group, with no statistically significant differences in height measurements between the treatment groups.
Hypothalamic-pituitary-adrenal axis function
Mean urine cortisol excretion corrected for creatinine (nanomoles/millimoles creatinine) was 7.9 for placebo (median, 6.3). The mean urine cortisol excretion corrected for creatinine excretion for ciclesonide treatment was 8.5 for 40 μg/d ciclesonide (median, 7.5), 7.0 for 80 μg/d ciclesonide (median, 5.5), and 9 for 160 μg/d ciclesonide (median, 8.44). The corresponding values for urine 6-β-hydroxycortisol were 39 for placebo (median, 35), 45 for 40 μg/d ciclesonide (median, 35), 33 for 80 μg/d ciclesonide (median, 31), and 43 for 160 μg/d ciclesonide (median, 32). No ciclesonide dose-related changes were seen in overnight excretion of creatinine-corrected urine cortisol or urine 6-β-hydroxycortisol (Table III).
Table III. Urine cortisol variables adjusted for creatinine: Differences between treatment groups
| Test – Reference∗ | ||||
|---|---|---|---|---|
| Test | Reference | LS Mean ± SE† | 95% CI | p value |
| Overnight 12-h urine cortisol (nmol/mmol creatinine) | ||||
| Placebo | 4.51 ± 9.35 | −14.1 to 23.1 | .63 | |
| Placebo | −1.44 ± 9.54 | −20.4 to 17.6 | .88 | |
| Placebo | 13.10 ± 9.46 | −5.7 to 31.9 | .17 | |
| Overnight 12-h urine 6-β-hydroxycortisol (nmol/mmol creatinine) | ||||
| Placebo | 6.22 ± 8.41 | −10.5 to 23.0 | .46 | |
| Placebo | −5.84 ± 8.59 | −22.9 to 11.3 | .50 | |
| Placebo | 5.09 ± 8.52 | −11.9 to 22.1 | .55 | |
∗Based on ANCOVA model with factors of treatment, sex, period, and patient (sequence) and covariates of age and baseline value. |
†Mean of the difference between the treatment and placebo values. |
Pulmonary function
Lung function was high during all treatment periods. The mean percent predicted FEV1 was 90.5% for placebo, 92.5% for 40 μg/d ciclesonide, 92.4% for 80 μg/d ciclesonide, and 91.9% for 160 μg/d ciclesonide. The median daytime or nighttime asthma symptom score was 0 during all treatment periods, including for placebo. No statistically significant variations were seen in these parameters or home diary recordings, including morning and evening peak expiratory flow, asthma symptom score, and rescue medication use, between the various treatments. The mean use of rescue medication during placebo was 0.2 puffs per day (range, 0-3.6 puffs per day).
Safety
No period or carryover effects or serious adverse events were seen. There were no cases of oropharyngeal candidiasis. The total number of adverse events reported during the 4 treatments was 5 during placebo, 7 during 40 μg/d ciclesonide, 4 during 80 μg/d ciclesonide, and 6 during 160 μg/d ciclesonide. Five of the adverse events were considered to be treatment related. One case of abdominal pain, cough, and rash occurred in a patient treated with placebo. One case of pharyngitis occurred in a patient treated with 80 μg/d ciclesonide, and 1 case of abdominal pain occurred in a patient treated with 160 μg/d ciclesonide.
Discussion
Studies have suggested that ciclesonide at 160 μg (exactuator dose, equivalent to 200 μg exvalve) per day is clinically effective in children and that, microgram for microgram, ciclesonide has at least the same clinical effect as fluticasone propionate.9 Similar to results for other ICSs, it has been difficult to show marked additional clinical improvements by increasing the ciclesonide dose to greater than 160 μg/d.8, 14 Therefore it was decided to study the effect of ciclesonide doses up to 160 μg/d on lower-leg growth rate and other markers of systemic ICS side effects (ie, cortisol suppression in children) in the present study. These clinically effective doses were found to be without any detectable effects on lower-leg growth rate or overnight urinary cortisol excretion.
Knemometry has been proved to be a very sensitive measurement of systemic effects of exogenous corticosteroids.12, 15, 16, 17, 18, 19 The growth-suppressive effects of ICSs on lower-leg growth rates are dose dependent, and the doses at which suppressive effects can be detected differ between different ICSs and different inhalation devicess.20 The same might be the case for the slopes of the growth-suppressive dose-response curves.20 The knemometry design used in the present study has previously been shown to be sufficiently sensitive to detect significant systemic effects of exogenous corticosteroids in a similar number of patients measured by the same investigators who participated in the present study.18 Thus 2.5 mg of prednisolone or 400 μg of beclomethasone diproprionate per day completely inhibited lower-leg growth rate over 2 weeks.12, 16 Furthermore, 400 μg of budesonide from Turbuhaler (AstraZeneca AB, Södertälje, Sweden), 400 μg of fluticasone propionate from Diskhaler (GlaxoSmithKline, Middlesex, United Kingdom), and 200 μg of mometasone furoate from a dry-powder inhaler significantly reduced lower-leg growth rate over 2 weeks.18, 21 Finally, in twice the number of patients monitored in the present study, mometasone furoate (100 and 200 μg/d) from a dry-powder inhaler demonstrated significant lower-leg growth-suppressive effects over 2 weeks.21 Therefore we do not believe that the lack of any adverse effects on lower-leg growth rates in the present study was due to insufficient power of the study to detect important differences.
The finding that 2.5 mg of prednisolone or 400 μg of beclomethasone diproprionate per day completely inhibited lower-leg growth rates, even if these treatments do not have such marked effects on statural growth,12, 16 indicate that knemometry is sensitive enough to detect any ICS-related growth effect that will be clinically important. In agreement with this, knemometry studies have always found significant suppression of lower-leg growth rates by ICS doses that do affect statural growth over 1 year22 and sometimes also significant effects of doses that do not have any adverse effects on statural growth.21 Therefore it seems that knemometry can be used as a sensitive tool to detect systemic effects of ICSs and also to define doses that are unlikely to be associated with any adverse effect on long-term statural growth, whereas a statistically significant lower-leg growth suppression is a poor predictor of long-term statural growth.22 Therefore the lack of any effects on lower-leg growth rate in the present study strongly suggests that long-term treatment with ciclesonide doses of up to 160 μg/d is unlikely to adversely affect long-term statural growth.
Earlier studies have found little or no correlation between steroid-induced changes in cortisol excretion in the urine and changes in lower-leg growth rates.17 However, in these studies knemometry has always been more sensitive than urinary cortisol excretion in detecting systemic effects or differences in systemic effects between different doses of exogenous steroids. Because there might be dissociation between adverse effects on lower-leg growth and the hypothalamic-pituitary-adrenal axis, overnight urine cortisol and 6-β-hydroxycortisol excretion also were included as a safety outcome. However, similar to results for lower-leg growth rate, no significant side effects were found in these parameters. This is in good agreement with the findings in adults, in whom no significant effects on urinary cortisol excretion were seen at daily ciclesonide doses of up to 1600 μg.23 For comparison, some studies have found significant reduction in urinary cortisol excretion with daily doses of 200 μg of fluticasone propionate (200 μg of budesonidehad no effect in the same study) and beclomethasone.24, 25, 26 Overnight cortisol excretion was preferred to 24-hour urinary cortisol excretion because it was more convenient, and its value has been validated in several trials.27, 28, 29
Our study group consisted of children with mild asthma in an effort to minimize any possible influence of poorly controlled asthma on growth. Furthermore, clinical effect parameters were measured to ensure that the asthmatic condition did not deteriorate to an extent that might interfere with normal lower-leg growth rate. The good lung function and the rare occurrence of symptoms in patients enrolled in the study confirmed the lack of asthma deterioration throughout the study. However, selection of patients with mild disease prohibited accurate measurement of treatment effects. Therefore it was not surprising that no differences were observed between active treatments and placebo because patients' symptoms were well controlled, and further improvements in the outcomes measured could not occur. Efficacy assessments would have required enrolling patients with more severe disease or including other outcome measures, such as bronchial hyperresponsiveness, exercise-induced asthma, or exhaled nitric oxide.
In the absence of detectable clinical effects, it is important to know that the lack of systemic effects was not due to poor compliance with the study medication. In this respect a correct inhalation technique was ensured repeatedly throughout the study. Moreover, the reported adherence (although admittedly not the best measure of adherence) was high and of the same magnitude as in other knemometry studies in our unit in which we have objectively measured adherence.18 Therefore we find it unlikely that the results in the present study were caused by poor adherence. More likely, they corroborated the low systemic effects with ciclesonide reported in adults.23
We thank Dr Christian Biberger and Dr Mechthild Barth (Altana Pharma AG, Konstanz, Germany) for logistical support.
References
- . Ciclesonide: an on-site-activated steroid. In: Hansel TT, Barnes PJ editor. New drug for asthma, allergy and COPD. Progress in respiratory research. Vol. 31:Basel (Switzerland): Karger; 2001;p. 91–93
- . Ciclesonide. Drugs of the Future. 2001;26:1033–1039
- . In vitro and in vivo anti-inflammatory activity of the new glucocorticoid ciclesonide. J Pharmacol Exp Ther. 2004;309:249–258
- . Pharmacokinetics of [14C]ciclesonide after oral and intravenous administration to healthy subjects. Clin Pharmacokinet. 2004;43:479–486
- . Ciclesonide (Byk Gulden). Curr Opin Investig Drugs. 2002;3:78–83
- . Risk-benefit value of inhaled glucocorticoids: a pharmacokinetic/pharmacodynamic perspective. J Clin Pharmacol. 2004;44:37–47
- Once daily ciclesonide is as effective as fluticasone propionate given twice daily in treating patients with asthma. [abstract] Am J Respir Crit Care Med. 2004;169(suppl):A91
- Maintenance of asthma control by once-daily inhaled ciclesonide in adults with persistent asthma. Allergy. 2005;60:330–337
- Pedersen S, Garcia ML, Manjra AL, Vermeulen JH, Therron I, Engelstätter R. Ciclesonide is as effective as fluticasone propionate in the treatment of children with persistent asthma. In: Proceedings of the 23rd Annual Congress of European Academy of Allergology and Immnonology; June 12–16, 2004; Amsterdam, the Netherlands.
- . Treatment of asthma by the inhaled corticosteroid ciclesonide given either in the morning or evening. Eur Respir J. 2001;17:1083–1088
- Once daily ciclesonide (80 μg or 320 μg) is equally effective as budesonide 200 μg given twice daily: a 12-week study in asthma patients. [abstract] Eur Respir J. 2003;22(suppl 45):410s
- . Short term linear growth in asthmatic children during treatment with prednisolone. BMJ. 1990;301:145–148
- . Accurate measurements of lower leg length and the ulnar length and its application in short term growth measurements. Growth. 1983;47:53–66
- Comparative study in asthma patients treated with inhaled ciclesonide (80 μg or 320 μg once daily) or budesonide (200 μg twice daily) for 12 weeks [abstract]. Am J Respir Crit Care Med. 2003;167(suppl):A771
- . Controlled study of linear growth in asthmatic children during treatment with inhaled glucocorticoids. Pediatrics. 1992;89:839–842
- . Short-term growth during treatment with inhaled fluticasone propionate and beclomethasone dipropionate. Arch Dis Child. 1993;68:673–676
- . Measures of systemic activity of inhaled glucocorticoids in children: a comparison of urine cortisol excretion and knemometry. Respir Med. 1995;89:347–349
- . Short-term Knemometry and urine cortisol excretion in children treated with fluticasone propionate and budesonide: a dose response study. Eur Respir J. 1997;10:1507–1512
- . Short-term lower leg growth rate in children with rhinitis treated with intranasal mometasone furoate and budesonide. J Allergy Clin Immunol. 1999;104:948–952
- . Do inhaled corticosteroids inhibit growth in children?. Am J Respir Crit Care Med. 2001;164:521–535
- . Lower-leg growth in children with asthma during treatment with inhaled corticosteroids. [abstract] J Allergy Clin Immunol. 2003;111(suppl):S269
- . Assessing the effect of intranasal steroids on growth. J Allergy Clin Immunol. 2001;108(suppl 1):40–44
- Drollmann A, Weinbrenner A, Hüneke D, Ziesche M, Engel G, Bethke T, et al. Repeated inhalation of the new topical steroid ciclesonide does not lead to suppression of endogenous cortisol release in healthy subjects [abstract]. Poster presented at: 11th Annual European Respiratory Society (ERS); September 22–26; Berlin, Germany.
- . Short-term knemometry and urine cortisol excretion in children treated with fluticasone propionate and budesonide: a dose response study. Eur Respir J. 1997;10:1507–1512
- Adrenal function in children with bronchial asthma treated with beclomethasone dipropionate or budesonide. J Allergy Clin Immunol. 1988;81:1088–1095
- . Twelve-month safety and efficacy of inhaled fluticasone propionate in children aged 1 to 3 years with recurrent wheeze. Pediatrics. 2004;113:e87–e94
- . A comparison of the systemic bioactivity of inhaled budesonide and fluticasone propionate in normal subjects. Br J Clin Pharmacol. 1994;38:527–532
- . Assessment of the systemic effects of inhaled glucocorticosteroids: the influence of blood sampling technique and frequency on plasma cortisol and leucocytes. Eur J Clin Pharmacol. 1990;39:127–131
- . 24-Hour and fractionated profiles of adrenocortical activity in asthmatic patients receiving inhaled and intranasal corticosteroids. Thorax. 1999;54:20–26
Supported by Altana Pharma AG, Konstanz, Germany.Disclosure of potential conflict of interest: L. Agertoft participated in advisory board meetings with AstraZeneca and GlaxoSmithKline. S. Pedersen within the last 3 years has participated in advisory board meetings with AstraZeneca, GlaxoSmithKline, Altana Pharma, and Merck Sharp & Dohme, and his institution has received grants from AstraZeneca, GlaxoSmithKline, Altana Pharma, Merck Sharp & Dohme, and 3M Pharmaceuticals.
PII: S0091-6749(05)00340-4
doi:10.1016/j.jaci.2005.01.066
© 2005 American Academy of Allergy, Asthma and Immunology. Published by Elsevier Inc. All rights reserved.
Volume 115, Issue 5 , Pages 940-945, May 2005
