Long-Acting beta2-Agonist Step-off in Patients With Controlled Asthma: Systematic Review With Meta-analysis
Brozek JL, Kraft M, Krishnan JA, et al
Arch Intern Med. 2012;172:1365-1375
The Study
When asthma has been controlled by adding a long-acting beta-2 agonist (LABA) to an inhaled steroid, as guidelines recommend, the US Food and Drug Administration (FDA) has recommended the LABA be withdrawn. This recommendation has raised concerns that the patient's asthma may relapse.
The authors of the current meta-analysis searched the literature to determine whether there was support for such discontinuation of LABA therapy. They screened 1492 publications for randomized controlled trials in which the LABA was withdrawn from use with inhaled corticosteroid (ICS) after restoration of asthma control.
Five published trials provided relevant information. Each was a prospective, randomized controlled trial in which adults with asthma that had been well controlled on a combination of an ICS and a LABA for at least 4 weeks were randomly assigned to discontinue the LABA. The ICS dose was unchanged, and short-acting bronchodilators were permitted on an as-needed basis. Observations of asthma control were made over the next 12 weeks or more.
Among patients in whom treatment with LABAs was discontinued, there were no statistically significant benefits and adverse outcomes were significantly more common. Thus, the results of the Asthma Control Test and Asthma Quality of Life Questionnaire were both worse; symptom-free days were fewer; and there was an increase in the use of rescue bronchodilators and a nonsignificant trend toward a need for a course of systemic corticosteroids. More than 3 times as many patients in the no-LABA group withdrew from the study owing to "lack of efficacy." No deaths or hospitalizations occurred in either group.
Viewpoint
Because these results were based on 5 studies in which the methodologies were not identical, there is some uncertainty about the validity of the conclusions, as there is with all meta-analyses. However, all of the data point toward loss of asthma control when the LABA was discontinued in a patient whose asthma had been well controlled by LABA/ICS combination therapy.
Concerns about the safety of LABAs in asthma were first expressed more than a decade ago, when a large postmarketing study of a LABA seemed to indicate an increased risk for mortality associated with one such agent. Although that study was of questionable design and the mortality effect was quite small, similar trends appeared in other studies, including of other LABAs. However, there were some indications that coadministration of a steroid with the LABA reduced the safety risk. Since then, monotherapy of asthma with a LABA has been advised against.
In 2010, the FDA issued a safety warning that recommended the use of a LABA only when the asthmatic patient was not well controlled with an ICS, and that the LABA should be discontinued when control had been reestablished. This has caused much uncertainty and concern among patients as well as health care providers.
Many providers questioned what constitutes "control" and how long control should be established before the LABA is withdrawn. In addition, is there not a risk for loss of control after LABA withdrawal? There was no guidance concerning these questions, and systematic studies designed to address the issue of LABA withdrawal are scant. Such studies as the current one are a first attempt to determine the wisdom of withdrawing the LABA after control of asthma has been reestablished.
The result of this meta-analysis suggests that patients whose asthma had been well controlled by the addition of a LABA to an ICS did worse when the LABA was subsequently withdrawn. Outcomes such as control of asthma, quality of life, and increased use of short-acting rescue bronchodilators were more common in the months after LABA discontinuation. Although these findings cannot be regarded as definitive, they may discourage healthcare providers from following the FDA's recommendation. In addition, the study does not address the issue of whether concomitant use of an ICS diminishes any risk of LABA use.
There is a need for further prospective long-term studies to address these questions. Hopefully, they will also throw light on the mechanisms by which a LABA might have an adverse effect.