KEYSTONE, COLO. — A meta-analysis of more than 23,000 asthma patients randomized to formoterol-containing regimens or to treatment without a long-acting beta-adrenergic agent showed zero asthma-related deaths and a consistent trend of fewer asthma-related hospitalizations in the formoterol/combo-treated patients.
The analysis of all 42 AstraZeneca-sponsored randomized, blinded clinical trials showed no evidence of increased risks of all-cause mortality, asthma-related deaths, or intubations in patients on combination therapy with the long-acting beta-agonist (LABA) formoterol.
Such findings support those of an earlier meta-analysis (Ann. Intern. Med. 2008;149:33-42) involving all of the more than 28,000 patients in the 66 GlaxoSmithKline-sponsored randomized trials comparing the LABA salmeterol plus an inhaled corticosteroid (ICS) versus an ICS alone, Dr. Harold S. Nelson said at a meeting on allergy and respiratory diseases.
These two meta-analyses paint a consistent picture of the safety of LABAs when used in conjunction with an ICS. It's a picture at odds with the “rather frightening” conclusions about LABA safety drawn by the Food and Drug Administration's Office of Surveillance and Epidemiology at a December 2008 meeting, said Dr. Nelson, professor of medicine at the University of Colorado/National Jewish Health, Denver.
The FDA analysis included only 1,270 of the 23,510 patients in the new meta-analysis. The regulators excluded data on non–U.S.-approved drug dosages and age groups, yielding a limited database from which they derived difficult-to-understand conclusions, the physician said at the meeting, sponsored by the National Jewish Medical and Research Center.
The FDA concerns about LABA safety arose in large part from the Salmeterol Multicenter Asthma Research Trial (SMART). But one by one, the major SMART conclusions—that salmeterol is associated with increased risk of asthma-related mortality, and that African Americans and children are uniquely vulnerable to LABA-related asthma exacerbations, as are patients homozygous for arginine at codon position 16 on the beta-2 adrenergic receptor—have subsequently been knocked down, said Dr. Nelson, who was the lead author of SMART (Chest 2006;129:15-26) but has been among those who have taken a wrecking ball to the study.
A major problem with SMART was that compliance with ICS therapy was not monitored, and apparently many patients assigned to salmeterol weren't taking the topical anti-inflammatory agent. “There's no question that the outcomes with combination therapy in SMART were bad. The only question is was it because they weren't taking an inhaled corticosteroid,” he said.
The results of the two large meta-analyses underscore the folly of much-publicized editorials calling for a new and supposedly definitive prospective trial of the safety of LABAs (Eur. Respir. J. 2009;33:3-5; N. Engl. J. Med. 2009;360:1671-2). “We've got more than 50,000 patients in clinical trials without an asthma death. … The number of patients that would be required for this new study would be somewhere between 1 million and infinity,” the physician said.
The new AstraZeneca-supported meta-analysis included 13,542 patients on formoterol-containing combination therapies and 9,968 on non-LABA regimens. The relative risk of asthma-related hospitalization was 27% lower in the formoterol-treated subjects. This difference didn't achieve significance because of the small number of hospitalizations, but Dr. Nelson noted that the trend in this and in the other outcomes consistently favored formoterol combination therapy. (See box.) Also noteworthy was the finding that the asthma hospitalization rate was similar in patients on daily doses of formoterol as low as 4.5 mcg and as high as 36 mcg or more.
Regarding other concerns raised by SMART, Dr. Nelson and his coworkers showed in a 544-patient, 16-week randomized trial that patients homozygous for arginine at codon position 16 of the beta-2 adrenergic receptor did not have worse morning peak expiratory flow than those with other genotypes. Measures of lung function were consistently more favorable with salmeterol plus fluticasone than with fluticasone alone (Am. J. Respir. Crit. Care Med. 2009 Nov. 12 [doi:10.1164/rccm.200809-1511OC]).
The supposedly greater susceptibility to asthma exacerbations seen among African Americans on LABA therapy in SMART hasn't been confirmed, either. A randomized trial involving 475 African Americans with asthma showed a mean annual asthma exacerbation rate of 0.449 cases per patient assigned to salmeterol plus fluticasone administered via the Advair Diskus device, compared with 0.529 cases per patient with fluticasone alone.
In children on formoterol and non-LABA regimens, the AstraZeneca meta-analysis showed similar asthma hospitalization rates. The same was true in the GlaxoSmithKline meta-analysis of salmeterol plus ICS versus ICS alone.
The FDA has ordered major changes to LABA product labeling, including a statement that LABAs are contraindicated without the concomitant use of asthma control medications such as inhaled corticosteroids. Dr. Nelson said that he has no quarrel with that, but he is concerned that the regulators' failure to grasp the benefits and safety of LABA/inhaled steroid combination therapy will not well serve patients and physicians. He predicted that LABA/ICS combinations are likely to become the treatment of choice as maintenance and reliever therapy for asthma, except in the United States.