Inhaled human insulin: new drug. No short-term advantages, too many unknowns in the long term.

(1) The standard treatment for type 1 diabetes is intensive insulin therapy, with at least 3 daily subcutaneous injections. Insulin is sometimes useful in type 2 diabetes, in which case the first-line treatment is an injection of isophane insulin at bedtime, in addition to ongoing oral antidiabetic therapy. (2) Pfizer has been granted marketing authorization in the EU for a powdered insulin product for pulmonary inhalation, for the treatment of adults with type 1 or type 2 diabetes. Two dose strengths are available (1 and 3 mg). (3) When inhaled, the insulin powder acts as rapidly as subcutaneous lispro insulin and lasts as long as a standard insulin injection. (4) Inhalation of 1 mg of insulin powder has similar glucose-lowering effects as 3 units of subcutaneous insulin, but inhalation of 3 mg is comparable to 8 units rather than 9 units of injected insulin. Three inhalations of 1 mg each have more glucose-lowering potency than a single inhalation of 3 mg. (5) None of the clinical trials published thus far have assessed the effects of inhaled insulin on clinical complications of diabetes. (6) In patients with type 1 diabetes, 7 randomised trials have compared inhaled insulin plus 1 or 2 subcutaneous injections of long-acting insulin with standard or intensive insulin therapy. They failed to show that intensive insulin therapy consisting of 3 insulin inhalations plus 1 or 2 injections of long-acting insulin reduced the HbA1c concentration or the frequency of hypoglycaemia more effectively than standard insulin therapy consisting of 2 daily subcutaneous insulin injections. (7) In type 2 diabetes, the addition of inhaled insulin has not been compared with the addition of injected insulin in patients whose glycaemia is not controlled by oral antidiabetic therapy. (8) In type 2 diabetes, 3 randomised trials have compared intensive insulin therapy based on inhaled insulin to subcutaneous insulin (2 to 4 daily injections), without oral antidiabetic drugs. The results suggest that glycaemic control is similar with both treatments. (9) The adverse effects of inhaled insulin have been assessed in fewer than 4000 patients participating in clinical trials, fewer than 600 of whom were treated for more than a year. During treatment lasting a few months, the most frequent short-term adverse effects (other than hypoglycaemia) seem to be mild respiratory adverse effects (cough, upper airway infections, etc.). (10) Treatment with inhaled insulin causes a gradual reduction in the peak expiratory flow rate (not convincingly shown to be reversible after the end of treatment) as well as a high incidence of anti-insulin antibodies. The possible long-term clinical consequences of these changes are unknown. The results of planned, long-term comparative trials should be available in 2014-2016. (11) The assessment of inhaled insulin in patients with respiratory disorders is inadequate. The effect of acute respiratory tract infections on the efficacy of inhaled insulin has not been adequately assessed. (12) Smoking (active or passive) and salbutamol, to a lesser extent, have important effects on the efficacy of inhaled insulin. (13) The insulin powder is very sensitive to high humidity, which can occur under normal conditions, leading to a risk of under-dosing. (14) The inhalation device is much larger than an injector pen. It does not permit precise insulin dose adjustment and delivers a maximum of 8 units per inhalation. (15) In practice, the many unknowns concerning the adverse effects of long-term treatment with inhaled insulin powder will probably not be resolved before 2016. In the meantime, subcutaneous injection remains the standard method of insulin delivery.