Background: In the United States, less than one half of adults with type 2 diabetes have an A1C level less than 7 percent. Despite this poor control, many patients resist taking subcutaneous insulin. Early in 2006, the U.S. Food and Drug Administration (FDA) approved a recombinant DNA inhaled insulin (Exubera) for the treatment of type 1 and type 2 diabetes in nonsmoking adults without pulmonary disease.

The Study: Ceglia and colleagues conducted a meta-analysis to evaluate the effectiveness of inhaled insulin compared with subcutaneous insulin and oral hypoglycemic agents. Randomized controlled trials of at least 12 weeks' duration that compared inhaled insulin with other therapies in adults with type 1 or 2 diabetes were included. The authors searched Medline and the FDA briefing on Exubera.

Results: The authors found 16 trials involving a total of 4,023 participants. There were two manufacturer studies that followed patients for two years, but most studies followed patients for 12 to 24 weeks. In trials that compared inhaled with subcutaneous insulin, subcutaneous insulin resulted in slightly lower A1C levels. There was no difference in the number of patients who achieved A1C levels of less than 7 percent. Overall, in patients with type 2 diabetes, inhaled insulin lowered A1C levels slightly more than oral hypoglycemic agents. However, the results wereheterogeneous. More patients with type 2 diabetes who used inhaled insulin had an A1C level less than 7 percent compared with those who used oral agents (30.9 versus 16.9 percent). Although more patients who used inhaled insulin reported at least one hypoglycemic episode than did those who used oral agents (9.4 versus 3.5 percent), there was no difference in incidence of hypoglycemia between the inhaled and subcutaneous insulin groups.

Inhaled insulin was associated with an increased risk of cough and a decrease in forced expiratory volume in one second (FEV₁). About 2 percent of patients discontinued inhaled insulin because of respiratory events compared with almost none in the comparison groups. The four trials that reported patient satisfaction showed that patients favored inhaled insulin.

Conclusion: Because inhaled insulin causes high insulin antibodies, cough, and decreased FEV₁ (even in nonsmokers with normal lung function), the authors conclude that there is reason to be cautious about offering patients with diabetes inhaled insulin as an alternative to oral agents or subcutaneous insulin. Long-term safety data on inhaled insulin are not yet available. There are no data on patients with diabetes who use inhaled insulin during upper or lower respiratory infections. The authors concede that none of the trials in their review are placebo controlled, and the acceptability of inhaled insulin may be related to its novelty. Cost-effectiveness studies have not been completed. The authors recommend postmarketing studies of inhaled insulin to refine the understanding of its place in the treatment of diabetes.