Compare Two Dosing Algorithms in Insulin-Naive Patients With Type 2 Diabetes Mellitus

This randomized, multicenter, open-label, active-comparator, 2-arm, parallel-group, 6 month
study with approximately 360 patients will compare the efficacy of two treatment regimens
(Algorithm A versus Algorithm B) in insulin-naive patients with type 2 diabetes not optimally
controlled by one or more oral antihyperglycemic medications. Patients will be assigned
randomly to receive one of the following treatment groups:

Algorithm A is defined as a simplified diabetes management regimen starting with a fixed dose
of HIIP (also known as AIR® Inhaled Insulin)(AIR® is a registered trademark of
Alkermes,Inc.), titrating 2 times per week based on 2 times per week 4 point blood glucose
values for the first month and titrating 1 time per week based on once weekly 4-point blood
glucose values for the remainder of the study, increasing total daily dose by a maximum of 6
U per day.

Algorithm B is defined as an intensive diabetes management regimen, starting with an adjusted
dose of AIR® Inhaled Insulin, titrating 2 times per week based on daily 4 point blood glucose
values, with sustained monitoring of dose and blood glucose throughout the study, increasing
total daily dose by a maximum of 8 U per day.