|
|
Background It is unknown whether inhaled corticosteroids can modify the subsequent development of asthma in preschool children at high risk for asthma.
Methods We randomly assigned 285 participants two or three years of age with a positive asthma predictive index to treatment with fluticasone propionate (at a dose of 88 µg twice daily) or masked placebo for two years, followed by a one-year period without study medication. The primary outcome was the proportion of episode-free days during the observation year.
Results During the observation year, no significant differences were seen between the two groups in the proportion of episode-free days, the number of exacerbations, or lung function. During the treatment period, as compared with placebo use, use of the inhaled corticosteroid was associated with a greater proportion of episode-free days (P=0.006) and a lower rate of exacerbations (P<0.001) and of supplementary use of controller medication (P<0.001). In the inhaled-corticosteroid group, as compared with the placebo group, the mean increase in height was 1.1 cm less at 24 months (P<0.001), but by the end of the trial, the height increase was 0.7 cm less (P=0.008). During treatment, the inhaled corticosteroid reduced symptoms and exacerbations but slowed growth, albeit temporarily and not progressively.
Conclusions In preschool children at high risk for asthma, two years of inhaled-corticosteroid therapy did not change the development of asthma symptoms or lung function during a third, treatment-free year. These findings do not provide support for a subsequent disease-modifying effect of inhaled corticosteroids after the treatment is discontinued. (ClinicalTrials.gov number, NCT00272441
[ClinicalTrials.gov]
.)
In school-age children, treatment with inhaled corticosteroids is associated with significant improvements in various measures of asthma control; these improvements disappear within months after the inhaled corticosteroids are discontinued.11,12,13,14 We hypothesized that the previous failure to observe a sustained effect on asthma after the discontinuation of inhaled corticosteroids was attributable to the possibility that the intervention was initiated too late in life.15 We thus designed the Prevention of Early Asthma in Kids (PEAK) clinical trial to determine whether the subsequent development of asthma in high-risk young children with frequent wheezing could be averted by starting inhaled corticosteroids early in life,16 before the development of persistent disease and chronic loss of lung function.15
Methods
A detailed description of the design of the PEAK trial, including screening and recruitment procedures and statistical analysis, has been reported in detail elsewhere.16 A description is also provided in the Methods section of the Supplementary Appendix (available with the full text of this article at www.nejm.org).
Screening and Schedule of Procedures
The PEAK trial was a multicenter, double-blind, randomized, placebo-controlled, parallel-group trial of inhaled fluticasone as compared with placebo in children two or three years of age who were at high risk for asthma (Figure 1A).16 Children were randomly assigned to receive either the inhaled corticosteroid or placebo for two years, at which point treatment was stopped. The children were then followed for an additional year, during which the primary outcome indicators were measured. The enrolled children had no clinically significant medical disorders apart from wheezing or allergy and were at high risk for the persistence of asthma-like symptoms during their school years, according to a positive modified asthma predictive index (Table 1 of the Supplementary Appendix).3,17 Only children who had received not more than four months of treatment with inhaled corticosteroids before enrollment and whose asthma symptoms did not require inhaled corticosteroids during a run-in month were eligible. After their parents provided written informed consent, a total of 285 children were randomly assigned, between January 2001 and January 2002, at five clinical centers (as described in the Supplementary Appendix). The schedule of procedures is shown in Figure 1A. Parents of the children were contacted by telephone to collect data on asthma-like symptoms (coughing and wheezing), asthma-medication use, and use of health care resources for respiratory symptoms during the preceding two weeks.
|
Treatments
The randomized treatment period ended in January 2004, and the observation phase ended in January 2005. Children were randomly assigned to one of two study groups: one that was to receive an inhaled corticosteroid (fluticasone propionate [Flovent] at a dose of two 44-µg puffs twice daily by metered-dose inhaler) and one that was to receive matching placebo (both provided by GlaxoSmithKline). These inhaled treatments were delivered by means of a valved spacer (AeroChamber) with a mask (provided by Monaghan Medical). Assignments were made with permuted-block randomization, with stratification according to clinic and age group.16 Adherence was promoted by an educational program16 and measured with use of an electronic meter (Doser). Rescue therapy in the form of two puffs of albuterol (90 µg per puff) or albuterol nebulization (2.5 mg) (both provided by Schering-Plough) was used according to a written action plan.16 A four-day course of oral prednisolone (provided by Muro Pharmaceutical) was prescribed according to the protocol for exacerbations, as described previously.16 The details of the protocols for adding or discontinuing supplementary medications, such as montelukast (Merck), and the criteria for the designation of treatment-failure status are described elsewhere16 and in the Supplementary Appendix. The PEAK study was extended, and the children continued to be observed through October 2005.
Outcome Measures
The primary outcome was the difference between the study groups in the proportion of episode-free days during the year-long observation period. Episode-free days were defined as those during which there were no asthma-like symptoms, no unscheduled medical visits for respiratory symptoms, and no use of any supplementary asthma medications, including albuterol before exercise. Episode-free days were reported by the parents during interviews and were based on two-week recall.18,19 Secondary outcomes included the proportion of episode-free days during the treatment period.
Other outcomes during both the treatment period and the observation period included the percentage of eosinophils and the number of courses of systemic corticosteroids and controller medications and the time until their use (Figure 1A). Impulse oscillometry, which assesses the contribution of resistance and reactance to the total impedance of the respiratory system, was performed according to established procedures (described in the Methods section of the Supplementary Appendix). Reactance at low frequencies reflects the elastic properties of the respiratory system and will have a lower magnitude in patients with airway dysfunction.20,21,22 Spirometry was attempted but was successfully completed in only 56 percent of the participants at the end of both the treatment and observation periods, and therefore, the results are not presented in this article. An exacerbation was defined as the need for a course of prednisolone to control asthma-like symptoms, as indicated in the protocol. In addition, potential side effects of long-term use of inhaled corticosteroids on growth were assessed. Height was measured at every visit with an upright stadiometer (Harpenden, Holtain) by established procedures.16,23 Skin-prick testing, with a core battery of 10 allergens in all the clinical centers and several locally relevant allergens in one center, was performed at enrollment and at the end of the treatment and observation periods16,17 (as described in the Methods section of the Supplementary Appendix).
Statistical Analysis
We determined that the enrollment of 280 children would provide the study with a statistical power of 90 percent, at the two-sided 0.05 significance level, to detect a difference between the groups of 10 percentage points in the proportion of episode-free days during the observation year. Primary statistical analyses were carried out according to the intention-to-treat principle. Study data, including data pertaining to growth and nonserious adverse events, were reviewed semiannually by a data and safety monitoring board (as described in the Supplementary Appendix).16 Serious adverse events were monitored and reviewed continually during the study; there were no formal, a priori stopping rules.
Episode-free days, exacerbations, and use of supplementary asthma medication were assessed by direct contact (by telephone or during a clinic visit) to evaluate asthma symptoms and medication use during the preceding 14 days and the use of health care resources since the last telephone call or clinic visit, as well as by review of the parents' written records of all medication used by the children. Episode-free days were determined from the parents' reported data, which were corrected according to the coordinators' records in cases in which the family did not report previously prescribed supplementary controller medication that had been recorded and dispensed by the coordinators. The proportion of episode-free days for each participant was calculated as the number of episode-free days divided by the number of days of observation. Data from all participants were used in the analysis regardless of the number of days observed.
Episode-free days during the treatment and observation periods were compared by an analysis of covariance (ANCOVA) in which the comparison between treatment groups was made after adjustment for age at randomization, sex, race or ethnic group (non-Hispanic white or other), center, aeroallergen skin-test reactivity (yes or no), duration of asthma-like symptoms at baseline (months), severity of symptoms at baseline (number of episode-free days during the run-in period), percentage of eosinophils, and eczema (yes or no). Episode-free days were converted to the logit scale before analysis in order to satisfy the distributional assumption associated with ANCOVA. For the post hoc analysis of responses, each child was characterized as either having had or not having had an apparent treatment response, separately for the observation year and for the two treatment years combined. These data were analyzed by contingency-table analysis and were summarized with the use of odds ratios. Participants were considered to have had an apparent treatment response if their proportion of episode-free days was greater than 92 percent, which was the median percentage of episode-free days in the group assigned to the inhaled corticosteroid during the two treatment years.
ANCOVA and contingency-table and regression analyses were carried out with use of SAS statistical software (version 8.2). All statistical tests were carried out at the two-sided 0.05 significance level. Additional details of all the analyses are provided in the Methods section of the Supplementary Appendix.
Results
Study Population
The two study groups were similar with respect to all baseline characteristics (Table 1), except that the group assigned to receive the inhaled corticosteroid had a higher percentage of peripheral-blood eosinophils. There were no significant differences between the groups in the numbers of completed clinic visits or telephone contacts, dropouts, treatment failures, or serious adverse events. Less than 12 percent of the participants in each group were lost to follow-up (i.e., dropped out) (Figure 1B). Children in whom treatment was deemed to have failed continued to be followed in the CARE clinics. Adherence to treatment, defined as the mean (±SD) percentage of days in which a child took the prescribed dose of study medication, as measured by an electronic meter, was similar in the two study groups (74±0.8 percent in the group assigned to inhaled fluticasone and 69±0.8 percent in the placebo group, P=0.10).
|
Observation Year
The proportion of episode-free days during the observation year, after discontinuation of the study drug, was not significantly different between the study groups, at 86.8 percent (95 percent confidence interval, 81.2 to 90.9 percent) in the group assigned to the inhaled corticosteroid and 85.9 percent (95 percent confidence interval, 79.9 to 90.3 percent) in the placebo group (P=0.78) (Figure 2). In addition, there were no significant differences between the study groups in the rate of use or the time to the initiation of systemic corticosteroid or montelukast treatment or in the rate of hospitalization.
|
|
During the two-year treatment period, the proportion of episode-free days was significantly greater in the fluticasone group than in the placebo group (93.2 percent [95 percent confidence interval, 91.1 to 94.9 percent] vs. 88.4 percent [95 percent confidence interval, 84.9 to 91.2 percent], P=0.006) (Figure 2). In addition, during the treatment period, the group assigned to inhaled corticosteroids had a lower rate of exacerbations necessitating a course of systemic corticosteroids than did the placebo group (57.4 per 100 child-years [95 percent confidence interval, 49.0 to 67.3] vs. 89.4 per 100 child-years [95 percent confidence interval, 78.3 to 102.2], P<0.001), less montelukast use (a mean of 11.4 days per year [95 percent confidence interval, 9.3 to 13.7] vs. a mean of 24.2 days per year [95 percent confidence interval, 20.5 to 27.9], P<0.001), and less supplementary use of fluticasone (i.e., use apart from that assigned in the study) (a mean of 8.3 days per year [95 percent confidence interval, 6.8 to 10.4] vs. a mean of 17.6 days per year [95 percent confidence interval, 14.9 to 20.9], P<0.001), but there was no significant difference between the groups in unscheduled visits to a physician or bronchodilator use (Table 2).
The time to the first course of systemic corticosteroids was also not significantly different between the study groups (P=0.26). However, the times to the second course of systemic corticosteroids (P=0.009 at 12 months) and to the initiation of treatment with montelukast (P=0.005), supplementary fluticasone (P=0.02), or other nonassigned asthma-controller medications (P=0.01) were significantly longer in the group assigned to inhaled corticosteroids than in the placebo group (Figure 3). No deaths, intubations, or hypoxic seizures during an exacerbation of asthma occurred in either group during the study.
|
In a post hoc analysis, more children in the group assigned to inhaled corticosteroids than in the placebo group were considered to have had an apparent response during the treatment period (proportion of episode-free days, >92 percent) (Table 3). In contrast, the proportion of children with a response who had a worsening (proportion of episode-free days, 
92 percent) during the observation year was significantly greater in the fluticasone group than in the placebo group. No significant difference between the treatment groups was seen during the observation year among the children who were considered not to have had a response during the treatment period (Table 3).
|
There was no significant difference between the study groups in any measure derived from impulse oscillometry at the end of the observation period (Table 2). In contrast, at the end of the treatment period, there was a significant reduction in the absolute magnitude of reactance at 5 Hz (–0.39±0.12 vs. –0.44±0.14 kPa per liter per second, P=0.008) and a trend toward a decrease in respiratory resistance at 5 Hz (1.01±0.20 vs. 1.05±0.20 kPa per liter per second, P=0.09) in the inhaled-corticosteroid group as compared with the placebo group (Table 2).
Measures of Growth
At the end of the observation year, the mean increase in height, relative to baseline, in the group assigned to inhaled fluticasone was 0.7 cm less than that in the placebo group (19.2±2.2 cm vs. 19.9±2.2 cm, P=0.008). This difference in the mean increase in height was smaller than the 1.1-cm difference observed at the end of the two-year treatment period (12.6±1.9 cm vs. 13.7±1.9 cm, P<0.001) (Figure 4). The difference in growth was detected during the first 12 months of inhaled-corticosteroid treatment as a difference in growth velocity between the study groups (6.6±1.0 cm per year in the fluticasone group vs. 7.3±1.0 cm per year in the placebo group between months 1 and 8, P=0.005) and a difference in the mean increase in height (4.5±1.1 vs. 4.9±1.1 cm, between months 4 and 12, P=0.001) (Figure 4). During the last 12 months of the treatment period, the growth velocity was similar in the study groups, and during the observation year the growth velocity in the inhaled-corticosteroid group was greater than that in the placebo group (7.0±0.8 cm per year vs. 6.4±0.9 cm per year from months 24 to 32, P=0.001) (Figure 4). Children in the inhaled-corticosteroid group had an average height percentile of 51.5±29.2 as compared with 56.4±27.3 among the children in the placebo group at the end of treatment (P<0.001) and a height percentile of 54.4±27.9 as compared with 56.4±26.9 at the end of observation (P=0.03).
|
In the PEAK trial, despite treatment of half the patients with an inhaled corticosteroid for two years, there was no effect on asthma-related outcomes during the observation year, after the corticosteroid treatment was stopped. The failure to observe a disease-modifying effect in the observation year cannot be attributed to an insufficient dose of fluticasone during the treatment period, since this agent affected both the burden of symptoms and somatic growth. Our findings indicate that these children had a considerable burden of symptoms, were at high risk for asthma, and had symptomatic benefit from daily therapy with an inhaled corticosteroid during the two-year treatment period. Like the study by Nielsen and Bisgaard,21 who analyzed impulse-oscillometric data in a similar population, the current study demonstrated that reactance at 5 Hz was less negative with the use of inhaled corticosteroids than with placebo. A lower absolute magnitude of respiratory-system reactance is a possible indicator of greater dynamic lung compliance,20,21,22 and therefore, this finding appears consistent with the other study outcomes, such as those pertaining to symptom burden, exacerbations, and supplementary-medication use.
Clinical improvement was observed while the children were treated with the inhaled corticosteroid but disappeared after treatment had been discontinued. Our data suggest that inhaled corticosteroids have little therapeutic effect on the processes that determine the progression of the disease from its initial, intermittent stages to a more chronic form, as described in the epidemiology literature.1,2,24,25,26,27
Elsewhere in this issue of the Journal and concordant with our findings, Bisgaard et al.28 report on the lack of effect of treatment with inhaled corticosteroids on the progression from episodic to persistent wheezing. In their study, no short-term benefit of episodic inhaled corticosteroids was demonstrated. This may be the result of the cohort's younger age, lower frequency of wheezing episodes before enrollment in the study, and the use of intermittent therapy three days after the onset of symptoms, rather than at the onset of symptoms.
On the basis of the burden of disease observed in the children enrolled in the trial, it appears that the use of a modified asthma predictive index was successful in identifying a group of children at high risk for continued asthma-like symptoms. In fact, these children had an increase in the frequency of asthma-like symptoms over time, and approximately 40 percent had an exacerbation that necessitated the use of oral corticosteroids during the observation year. The PEAK cohort was younger and had less frequent symptoms than the 5-to-12-year-old children with mild-to-moderate persistent asthma who were enrolled in the Childhood Asthma Management Program Research study,12 but the atopic characteristics (personal and family histories of atopy, food and aeroallergen sensitization, and eosinophilia)16 were similar in the two populations.
One might speculate that if therapy had been initiated before two years of age, the natural course of asthma might have been changed. However, starting long-term treatment without a valid predictive index in children with recurrent wheezing who are not yet two years of age would mean the unnecessary treatment of many children who will eventually outgrow their disease (i.e., those who wheeze transiently)15,29 and thus would be ethically questionable.
Moreover, one could speculate that the children who had a favorable response to inhaled corticosteroids during the treatment period would constitute the group most likely to have a sustained benefit during the observation year. In fact, post hoc analysis suggests that the opposite was true. Although more children in the inhaled-corticosteroid group than in the placebo group had a response during the treatment period, a greater proportion of them had worsening during the observation year, after the study medication was discontinued — a pattern similar to that in adults and older children with persistent asthma.11,13,14 These data do not imply that inhaled corticosteroids worsened the children's condition, since the two treatment groups had a similar burden of disease during the observation year.
A reduction in growth velocity was observed in the group assigned to inhaled corticosteroids during the first year of treatment. This observation differs from the normal growth velocities reported in a study of fluticasone treatment for one year5 among preschool children with recurrent wheezing. In that study, the use of a larger volume spacer for medication delivery,5,30 which may have led to a different corticosteroid-deposition pattern or to a lower adherence or higher dropout rate,5 may explain the difference in our findings. Although the observed growth rate in the placebo group was similar to that in the inhaled-corticosteroid group during the last 12 months of treatment, and although growth accelerated during the observation year, there remained a difference of 0.7 cm in height between the treatment groups at the end of the trial. It remains to be determined whether height will become similar in the two groups as the cohort matures, in a manner similar to that observed in older children.31
Our data show that the natural course of asthma in young children at high risk for subsequent asthma is not modified by two years of treatment with inhaled corticosteroids. The treatment, however, did reduce the burden of illness. Our findings demonstrate that inhaled corticosteroids can be used to control active disease but should not be used to prevent asthma in high-risk preschool children.
Supported by grants (HL071742-01, HL004519-04, 5U10HL064287, 5U10HL064288, 5U10HL064295, 5U10HL064307, 5U10HL064305, and 5U10HL064313) from the National Heart, Lung, and Blood Institute. This study was carried out in part in the General Clinical Research Centers at Washington University School of Medicine (M01 RR00036) sponsored by the National Institutes of Health and the National Jewish Medical and Research Center (M01 RR00051).
Dr. Guilbert reports having served on the advisory boards of GlaxoSmithKline and Innovia Educational Institute; having received lecture fees from GlaxoSmithKline, AstraZeneca, SOMA Medical Education, Innovia Educational Institute, and Antidote (formerly Medical World Conferences); having served as a subinvestigator for a research grant received from Genentech; and having received consulting fees for a research study from GlaxoSmithKline. Dr. Morgan reports having served on the advisory boards of Genentech and the Cystic Fibrosis Foundation. Dr. Zeiger reports having served on the advisory boards of Aerocrine, AstraZeneca, Genentech, GlaxoSmithKline, and Novartis and having received lecture fees from AstraZeneca and reports that his research group has received research support from GlaxoSmithKline and Sanofi-Aventis. Dr. Szefler reports having served on the advisory boards of AstraZeneca, GlaxoSmithKline, Aventis, Genentech, and Merck and having received research grants from Ross Pharmaceuticals and AstraZeneca. Dr. Bacharier reports having received lecture fees from GlaxoSmithKline, Merck, Genentech, and AstraZeneca. Dr. Lemanske reports having received consulting fees from AstraZeneca, Aventis, Novartis, GlaxoSmithKline, and Altana; having received lecture fees from Merck, GlaxoSmithKline, AstraZeneca, and Aventis; and having served as a subinvestigator in studies supported by research grants from Dey, Wyeth Pharmaceuticals, Pfizer, GlaxoSmithKline, MediciNova, Dynavax Technologies, Greer Laboratories, Aventis, Eli Lilly, Boehringer Ingelheim Pharmaceuticals, Merck, Centocor, Ono Pharmaceuticals, Altana, the University of Nebraska Food Allergy Research and Resource Program, Agennix, Atherogenetics, Kling Pharmaceuticals Research and Development, Abaris, Schering-Plough, and Fetzer. Dr. Allen reports having received lecture fees from GlaxoSmithKline. Dr. Krawiec reports having received lecture fees from Merck, Innovia Educational Institute, and GlaxoSmithKline and having received a research grant from Pfizer. Dr. Larsen reports having served on the advisory boards of GlaxoSmithKline and Schering-Plough. Dr. Liu reports having served on the advisory board of GlaxoSmithKline and having received lecture fees from GlaxoSmithKline, Merck, and AstraZeneca. Dr. Sorkness reports having served on the advisory boards of GlaxoSmithKline and AstraZeneca; having received lecture fees from GlaxoSmithKline, AstraZeneca, and Genentech; and having received a research grant from GlaxoSmithKline. Dr. Taussig reports having served on an advisory panel for GlaxoSmithKline. Dr. Martinez reports having served on the advisory board of Merck, having received lecture fees from Merck, having received consulting fees from Pfizer and Genentech, and being a member of the speakers' bureau of Merck. No other potential conflict of interest relevant to this article was reported.
Source Information
From the Arizona Respiratory Center, University of Arizona, Tucson (T.W.G., W.J.M., F.D.M.); Kaiser Permanente, San Diego, Calif. (R.S.Z.); the University of California–San Diego, San Diego (R.S.Z., G.H.); Pennsylvania State University, Hershey (D.T.M., S.J.B., V.M.C.); National Jewish Medical and Research Center, Denver (S.J.S., G.L., A.H.L., L.M.T.); Washington University, St. Louis (L.B.B., R.C.S., G.R.B.); and the University of Wisconsin, Madison (R.F.L., D.B.A., M.K., C.A.S.).
Address reprint requests to Dr. Guilbert at the Division of Pediatric Pulmonary Medicine, Arizona Respiratory Center, University of Arizona, 1501 N. Campbell Ave., P.O. Box 245073, Tucson, AZ 85724, or at guilbert{at}arc.arizona.edu.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Inhaled Corticosteroids and Children
Köhler D., Haidl P., Dellweg D., Bont L., Kimpen J. L.L., Ermers M. J.J., Baraldi E., Filippone M., Guilbert T., Martinez F. D., Szefler S. J., the Childhood Asthma Research and Education (CARE) Network Steering Committee , Bisgaard H., Gold D. R., Fuhlbrigge A. L.
Extract |
Full Text |
PDF
N Engl J Med 2006;
355:624-626, Aug 10, 2006.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
