The long-term benefits of glycemic control with intensive insulin therapy in patients with diabetes has only recently been demonstrated. Intensive therapy usually requires frequent injections of regular insulin. Multiple daily injections can be inconvenient and require substantial time, effort and communication for the patient and physician. Poor adherence to therapy is often the result. An alternate way to deliver insulin could use the lung, which is a highly permeable port of entry for macromolecules into the blood. Cefalu and associates conducted a randomized, open-label study to investigate the efficacy and safety of inhaled insulin in patients with type 2 diabetes mellitus.

Patients 35 to 65 years of age who were on a stable insulin regimen requiring two to three injections daily were enrolled in the study. Participants had no evidence of major organ system disease, and did not smoke or use concomitant oral antidiabetic agents. They were initially hospitalized for two days and instructed in the self-administration of inhaled aerosolized insulin using dry powder formulations in 1- or 3-mg doses. Inhaled insulin was used before each meal in one to two inhalations of the appropriate dose strength; their sole long-acting insulin was ultralente, which was used at bedtime. Each milligram of inhaled insulin delivers to the circulation the rough equivalent of 3 U of subcutaneous insulin. Dose titration of both the inhaled insulin and the bedtime subcutaneous dose was used to achieve a mean preprandial glucose value of 100 to 160 mg per dL (5.6 to 8.9 mmol per L).

Inhaled insulin significantly improved glycemic control as assessed by glycosylated hemoglobin (HbA_1c) levels in the 26 participants. Mild to moderate hypoglycemic events were experienced by 18 patients and no severe events occurred. There were no adverse pulmonary effects in terms of spirometry results, lung volume, diffusion capacity or oxygen saturation.

The authors conclude that inhaled insulin appears to have an earlier onset of action than subcutaneous insulin. They found that 1 mg of inhaled insulin has a peak level similar to approximately 3 U of insulin, allowing calculation of initial dosing. Glycemic control can be obtained with a combination of inhaled insulin and a single nighttime subcutaneous dose of long-acting insulin.

In an editorial in the same issue, Nathan points out that the decrease in HbA_1c levels seen in this study was disappointingly modest. According to Nathan, relatively large doses of inhaled insulin were required because of inefficient absorption from the alveoli. Perhaps the more important question in achieving good glycemic control is provision of adequate doses of insulin (usually 0.8 U per kg or more of body weight) rather than frequency of administration. If ambient plasma glucose levels are reduced, especially if this reduction is achieved in the initial stages of insulin resistance, functional beta cells may recover sufficiently to control daytime blood glucose. Nathan concludes that frequent injection, or inhaled therapy, may be optimal therapy in patients with type 1 diabetes. Better therapy in patients with type 2 diabetes would include higher insulin dosing and the initiation of insulin therapy earlier in the course of insulin resistance.