Do long-acting beta2-adrenergic agonists deserve a different place in guidelines for the treatment of asthma and COPD?

The introduction of long-acting b2-adrenergic agonists (salmeterol and formoterol) in the early 1990s came at a rather critical moment. Various epidemiological studies had recently demonstrated a relationship between prolonged use of b2-adrenergic agonists and mortality due to asthma [1±5]. These publications did not go unheeded: they triggered intense debate on the possible risks of the use of b2-agonists. One of the main criticisms that these studies elicited was that they did not provide any hard evidence for a cause/effect relationship. In other words, it could not be deduced from these studies that prolonged use of b2-agonists (fenoterol in particular) was the true cause of increased mortality due to asthma. Severe asthma is accompanied by not only higher mortality but also frequent use of b2-agonists. The connection between prescription of these agonists and mortality could therefore also be reversed (so-called "confounding by indication"). A meta-analysis based on the most important epidemiological studies confirmed the suspicion that there indeed existed a reverse relationship [6]. However, as meta-analysis of such patient/control studies always has a number of methodological limitations, there was a clear need for randomized prospective studies. In 1990/1991, two randomized longitudinal intervention studies were published, in which prolonged use of b2-agonists seemed to be the cause of diminished control over asthma [7] and a relatively rapid decline in lung function [8]. Partly because of this, there was growing concern regarding the possible harmful effects of prolonged use of these drugs. Although these findings were not supported by other studies [9±12], physicians (and patients) had the feeling that the recently introduced long-acting b2-agonists should be used with due caution. Further studies were therefore required. It turned out that prolonged use of not only short-acting b2-agonists, but also a long-acting agonist led to a decrease in the protective effects against bronchoconstrictive stimuli in the long term [13, 14]. However, this was not accompanied by an increase in bronchial hyperresponsiveness or a diminished bronchodilatory effect of these long-acting drugs [14, 15]. Although, at first sight, this last aspect seems positive, it simultaneously reinforces the idea that these very effective bronchodilators may mask the severity of the disease in the long run. After all, the long-acting b2-adrenergic agonists have not been shown to have a broad anti-inflammatory effect. Significant exposure to irritants (e.g. allergens or smoke) might unexpectedly cause a serious obstruction, with the patient not noticing any deterioration in their condition at the time, because all symptoms are suppressed so effectively by these long-acting drugs. Moreover, there is a risk that the absence of symptoms will lead to less compliance with regard to inhaled corticosteroids, which have just the antiinflammatory effect that b2-agonists do not have.