Three common asthma inhalers containing the drugs salmeterol or formoterol may be causing four out of five U.S. asthma-related deaths per year and should be taken off the market, researchers from Cornell and Stanford universities have concluded after a search of medical literature.

They base these conclusions on a statistical analysis of 19 published trials involving 33,826 patients. This so-called meta-analysis found that patients who inhaled the long-acting beta-agonists salmeterol (trade names Serevent and Advair, both made by GlaxoSmithKline) or formoterol (trade name Foradil, made by Novartis Pharmaceuticals) were 3.5 times more likely to die from asthma and 2.5 times more likely to be hospitalized (whether or not death resulted), compared with those taking a placebo.

The reason, say the researchers, is because although these medications relieve asthma symptoms, they also promote bronchial inflammation and sensitivity without warning.

Nevertheless, asthma death is relatively rare - 15 patients in the meta-analysis who were taking the beta-agonists died, compared with three in the placebo group, over a six-month period.

“In total, there are about 5,000 deaths a year due to asthma, whether or not a person is taking a long-acting beta-agonist,” said Edwin Salpeter, the J.G. White Distinguished Professor of Physical Sciences Emeritus at Cornell, who led the statistical analysis in the study. An eminent astrophysicist, Salpeter has more recently focused his attention on medical statistics. “We can show that overall it is statistically significant that, compared to patients taking a placebo, these long-acting beta-agonists kill a lot of people,” he said.

“These asthma deaths are generally in healthy young adults,” said his daughter, Shelley Salpeter, the lead author of the paper appearing online and in the June 20 issue of Annals of Internal Medicine. She is a clinical professor of medicine at Stanford’s School of Medicine and a physician at Santa Clara Valley Medical Center in San Jose, Calif.

“We estimate that approximately 4,000 out of the 5,000 asthma deaths that occur in the U.S. each year are actually caused by these long-acting beta-agonists, and we urge that these agents be taken off the market,” she added.

Adding an anti-inflammatory drug to a long-acting beta-agonist adds little benefit, report the Salpeters. Advair, for example, (the fourth bestselling drug in the world with $5.6 billion in annual sales), combines salmeterol with an anti-inflammatory drug to provide some protection against bronchial inflammation associated with beta-agonists. But hospitalizations still doubled for patients inhaling a long-acting beta-agonist combined with an anti-inflammatory drug compared with asthma patients taking a placebo and an anti-inflammatory drug by itself.

Of the 19 studies surveyed in the meta-analysis, the largest - the Salmeterol Multicenter Asthma Research Trial with 26,000 participants - reported a fourfold increased risk for asthma-related deaths and a twofold increase in life-threatening asthma events in patients using salmeterol. If older people who also suffered from chronic obstructive pulmonary disease were removed from the analysis, the Salpeters report, salmeterol would be associated with a six times greater risk for asthma-related deaths.

The meta-analysis found that 53 of 3,083 patients inhaling beta-agonists were hospitalized for an asthma attack compared with 12 of 2,008 patients who received a placebo, meaning that there was one hospitalization for every 71 patients treated with a long-acting beta-agonist per year.

The Salpeters say that these two long-acting beta-agonists can result in death because tolerance to them develops over time.

“These agents can improve symptoms through bronchodilation at the same time as increasing underlying inflammation and bronchial hyper-responsiveness, thus worsening asthma control without any warning of increased symptoms,” said Shelley Salpeter.

“It is particularly frightening that long-acting beta-agonists are detrimental, whether salmeterol or formoterol, whether taken alone or with anti-inflammatory drugs, both for children and for adults,” said Ed Salpeter.

Co-authors of the study include Shelley Salpeter’s son, Nicholas Buckley, a student at Sequoia High School in Redwood City, Calif., and Thomas Ormiston, a physician at Santa Clara Valley Medical Center in San Jose, Calif.