© 2004 American Thoracic Society
Progression of AsthmaSmall Steps and a Long Way to GoUniversity of Southern Denmark Kolding, Denmark Airway inflammation and remodeling are present in children and adults with asthma of every severity. Are these phenomena linked to long-term clinical outcomes of asthma, such as progressive loss of lung function and progression of clinical disease in some or all patients? Can their progression be modified by treatment or other interventions? These questions are addressed in this issue of the Journal by Covar and coworkers (pp. 234–241) (1) in a post-hoc analysis of individual declines in lung functions of 990 patients who participated in a long-term asthma trial (2–4). With respect to progressive loss of lung function, Covar and coworkers found a small, non-significant 0.9% decline in post-bronchodilator FEV1% predicted over 4 years for the whole population (1), whereas post-bronchodilator FEV1/FVC ratios decreased by 1.7% in the placebo group (2). At first glance, this finding suggests that decline in lung function is not a problem in children with mild asthma. This generalization, however, may be inappropriate and an oversimplification. In 7,000 patients (3,000 between 5 and 15 years) with newly diagnosed, mild asthma, Pauwels and coworkers (5) found the following 3-year declines in post-bronchodilator FEV1 in the placebo group: 2.3% in 6- to 10-year-olds, 0.4% in 11- to 18-year-olds, and 3.6% in an adult group. The declines were highly significant in the younger and older age groups, but not in the 11- to 18-year-olds. Other investigators have also reported significant declines in both pre- and post-bronchodilator lung functions in children with mild and moderate asthma treated with salmeterol (6) and placebo (7). In the latter study the decline seemed more marked in lung function parameters reflecting the more peripheral airways. Moreover, several cohort studies and cross-sectional studies in children and adults suggest that over time, groups of patients with asthma lose lung function at a greater rate than do subjects without asthma (4, 8–10). This decline may be more marked early in the disease (11) and in patients with more severe asthma (8, 12). The discrepancy between these findings and those reported by Covar and coworkers (1) is not known. Close examination of their data suggests two possible explanations. One, differences in asthma duration: Covar and coworkers (1) found that short duration of asthma was associated with greater decline in lung function. All patients in the study by Pauwels and coworkers (5) had a very short duration of asthma. This was not the case for the patients of Covar and coworkers. Two, differences in age: The mean age at study entry of the patients in the analysis by Covar and coworkers (1) was 9.5 years. So age-wise their patients resembled the 11- to 17-year-olds in the study by Pauwels and coworkers in whom no significant decline in post-bronchodilator FEV1 was seen (5). The finding by Covar and coworkers (1) that young age (< 9 years) was associated with increased decline in lung function supports this explanation. It is not known why the findings in adolescents should be different from other age groups. During puberty, the relationship between height and lung functions is more complex than in other periods of life and assessment of FEV1% predicted may not be as suitable a measure as in other ages (13, 14). So, more studies are needed in pubertal children before conclusions from these age groups can be generalized to other age groups. It is difficult to assess a treatment effect on decline in lung function when no significant decline is seen in the placebo-treated group. In an attempt to overcome this problem, Covar and coworkers (1) pooled the changes in post-bronchodilator FEV1 from all three treatment groups in the original study. They found that a substantial number of patients did not have any decline in lung function over the 4 years. This is new and important information, which may speak against inflammation and remodeling as the main cause of decline. Decline was seen only in approximately 60% and a "rapid decline" (defined as a decline of at least 1% per year) in 26% of the patients. Treatment did not influence the frequency of "rapid decliners." At first glance this could suggest that treatment does not modify decline in lung function. Selection, however, of a group of patients on one of the main outcomes in the original study is not without problems, and there are many reasons, why such post-hoc selection based on a rather arbitrary cut-off limit does not allow any firm additional conclusions about treatment effects. This must be obtained in prospective, placebo-controlled studies designed to answer this question. Such studies do suggest significant treatment effects. In the study by Pauwels and coworkers, 200 g budesonide per day significantly reduced the annual decline in post-bronchodilator FEV1% predicted by 22% in 2,000 young children, and 400 g per day reduced the decline by 42% in adults (both p < 0.001) (5). No treatment effect was seen in the adolescents whose FEV1 did not decline. In agreement with this observation, 600 g budesonide per day (but not placebo) was found to normalize post-bronchodilator lung function development of central and intermediate airways over two years (7). Small airway caliber also improved significantly, although it still remained somewhat reduced. In the patients analyzed by Covar and coworkers (1), the only variable (FEV1/FVC ratio) that declined somewhat over time was also the variable that came closest to a treatment effect (p = 0.08) (2). So treatment does seem to modify decline. More information, however, is needed about optimal doses, outcomes, and patients at risk. In this respect, the findings by Covar and coworkers (1) that young age, early diagnosis of asthma, male sex, short duration of asthma, and more prominent eosinophilic airway inflammation were more common in "rapid decliners" is useful. It must be remembered that FEV1 may remain normal even in the presence of marked damage to the smaller airways, so data on FEV1 alone does not exclude the possibility of clinically important changes in the peripheral airways. Future studies must therefore also include outcomes which reflect peripheral airway function. Covar and coworkers (1) agree with most researchers that post-bronchodilator lung-function values are preferable to pre-bronchodilator values, although this has not been formerly validated. Genetic heterogeneity of bronchodilator response or increased bronchodilator response over time as a result of remodeling of the airway smooth muscles might complicate the interpretation of post-bronchodilator values, particularly if treatment prevents this effect on airway smooth muscle. Thus, continuous treatment with inhaled corticosteroids is often associated with a diminished bronchodilator response over time, whereas placebo treatment is not (5, 7). FOOTNOTES Conflict of Interest Statement: S.P. has received €13,000 in 2001 and €12,000 in 2002 for speaking at scientific meetings or courses organized and financed by AstraZeneca and €8,000 in 2002 for speaking at scientific meetings organized and financed by GlaxoSmithKline. University of Southern Denmark has received unrestricted grants from AstraZeneca (€14,000 in 2001) and GlaxoSmithKline (€120,000 in 2001, 2002 and 2003). REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
