Asthma variability

To the Editor:

The article by Zhang et al. 1 demonstrated the ability to reanalyse previously published data and reformulate the conclusions. The analysis describes the hetereogeniety of asthma­treatment response among patients treated with inhaled corticosteroids, montelukast or placebo. Although these studies demonstrated that both active therapies are superior to placebo and that beclomethasone is superior to montelukast, appropriately, the authors point out that the response to specific end­points is inconsistent. They illustrate the response to active treatment quite well with distribution curves and conclude that there is no simple haplotype to describe a responder. They propose that response to therapy is dependent on multiple variables. There is a normal unimodal distribution curve for lung function and symptom scores for both active treatments as well as placebo. The active therapies show the distribution curve moving in the direction of improvement, with beclomethasone greater than montelukast. Studying the placebo response illustrates the variable nature of asthma, as some placebo­treated patients exhibit significant improvement, some significant worsening and many remain unchanged. Beyond studying the impact of the medication is the importance of understanding the variable nature of asthma, which is essential in understanding the true response or lack of improvement following therapy. The logic of finding a genetic response would also suggest that if one builds a model for studying pharmacogenomic haplotypes for active therapies, one must also consider a model to study placebo haplotypes.

The “within­patient variability” probably reflects the unpredictability of the disease. The definition of asthma of the National Heart, Lung and Blood Institute (NHLBI) 2 includes “widespread but variable airflow obstruction that is often reversible either spontaneously or with treatment.” This variable nature of the disease may be more important than finding the “surrogate clinical marker of asthma control.” Mean data may be relevant and studying multiple outcomes is essential. When the majority of end­points demonstrate that one therapy is better than the other, then the search for the ultimate end­point may be unnecessary. Although there is not a clear relationship between the end­points that are measured, what is unambiguous is that inhaled corticosteroids for all clinical end­points are superior to leukotriene­receptor anatagonists. This article is a good intellectual exercise but does not change previous conclusions. The ultimate end­point in evaluating asthma is to achieve the goals of asthma care: reduce symptoms and exacerbations. NHLBI 2 and the Cochrane Library Database 3 have used formal evidence­based criteria to review the literature and have concluded that inhaled corticosteroids are the preferred initial controller therapy for persistent asthma.

Rather than trying to re­evaluate data, new studies should be designed with alternative therapies that provide greater improvement in lung function, better symptom control and fewer exacerbations, with the understanding that the differences in response should also take into consideration the variable nature of the asthma.