Home / Resources / Articles / Inhaled Vs. Injected Insulin

Inhaled Vs. Injected Insulin

Oct 20, 2018
Editor: Joy Pape, MSN, FNP-C, CDE, WOCN, CFCN, FAADE

Author: Arsalan Hashmi, PharmD. Candidate, LECOM College of Pharmacy

A study tests Technosphere insulin, postprandial inhaled glucose for a more natural insulin effect.

Despite today’s advances in diabetes medicine and technology, many people who have type 1 diabetes still have trouble managing their glucose levels.  This is partly to blame on the mealtime insulins like Novolog, which may show peak levels up to 3 hours after injectio

Due to multiple physiological factors after injection of rapid-acting insulins, glucose levels may fluctuate from hyperglycemia to delayed hypoglycemia. A new technology is looking to solve this problem. Technosphere insulin (TI) is a dry-powder insulin that is inhaled, starts working in 12 minutes, and peaks in 35 to 45 minutes.

There have been previous studies that have shown noninferiority to Novolog, less weight gain, less hypoglycemic events, and even a reduction in A1C. This study was a four-week multicenter randomized clinical trial investigating the effect of additional postprandial TI inhalations versus a solely before-meal aspart injection (Novolog). The study utilized continuous glucose monitoring, time-in-range, prandial glucose excursions (difference between highest CGM glucose and premeal glucose), and time in hypoglycemia.


Inclusion Criteria Exclusion Criteria
Nonsmoking Pregnancy
BMI no greater than 35 kg/m2 Pulmonary diseases
FEV1 at least 70% Severe hypoglycemia in last 3 months
Stable insulin dose for minimum 3 months Smoking or marijuana
Basal insulin of glargine or degludec Basal insulin of detemir or NPH
T1DM HbA1c 6.5%-10.0% for at least 6 mo Needing >18 U prandial insulin

Initially all patients received an EKG, physical exam, CMP, and spirometry test. FEV1 was evaluated at baseline, and weeks 2 and 4 in both groups. Patients had seven clinic/phone visits, and were randomized 1:1 to TI or insulin aspart groups at the second visit. Patients were given Dexcom G5 real-time CGM for the study. Patients randomized to the aspart group were all converted to aspart insulin, and were allowed to change pre-meal bolus and take postprandial and correction doses if needed. Patients randomized to the TI group were given conversion tables on how to dose the inhalations.

Additionally, they were instructed to take extra inhalations at one and two hours after eating based on their postprandial glucose. Throughout the study, patients did not change their basal insulin dose. Primary endpoints included glucose time in range and postprandial glucose excursion from one to four hours. Secondary endpoints were HbA1c, AUC for postprandial glucose, hypo and hyperglycemia and glucose variability. Adherence  with the protocol for TI group was measured since participants in the group had to take two post-meal doses. Compliance was set to 90%, and at least one inhalation after meals. A student’s t-test was used to compare baseline characteristics, and linear regression models to examine CGM data.

Of the 60 participants randomized, four were not included in the final analysis. There were no differences in baseline and end FEV1. There was no severe hypoglycemia during the study. A mild cough was reported in the TI group by three participants. The TI group had much higher daily insulin bolus doses, but similar doses before meals. Patients in the TI group took both one- and two-hour post-meal TI inhalations 27% of the time. The aspart group took less extra doses when compared to the TI group. Even with the increased doses of TI, patients did not gain weight; instead slight weight loss was seen. Patients in the aspart group did gain weight.

The linear regression models had 22 patients from the TI group and 34 from the aspart group with adjustments made for many factors. Glucose fluctuation and time in hypoglycemia was significantly less in the TI group. The 15 patients who were adhered to the  the TI regimen had significantly higher time in range, lower glucose variation, and less hyperglycemia than both the control group and non-compliant group. Those who followed protocol experienced significantly less PP glucose excursions at breakfast and lunch, but not dinner. Variation in glucose was also significantly lower in the TI group who were not following the regimen during the daytime, but there was no found difference at night. TI patients who did adhere to the regimen also experienced significantly lower PPG up until 120 minutes after meals, with no glucose difference before meals.

This study showed overall improvements in all areas tested, except post-dinner PPGE. A possible explanation for this may be a patient’s previous experience with nocturnal hypoglycemia from longer acting mealtime insulins. Thus, they would not follow the correct regimen at night in fear of hypoglycemia; additional patient education is needed. The high amount of  not adherence may be due to the recommended dose being 1.5x higher than the subcutaneous dose.

The study did have limitations. The A1c for the aspart group was slightly higher than the TI group. There was no “run-in” period where the insulin doses were optimized. Patients wearing the continuous glucose monitors for the first time may have inadvertently changed a variable on the monitors without realizing it. The study could be underpowered with only 60 participants. The study was funded by MannKind, the company that makes the TI product.

Practice Pearls:

  • Patients on TI therapy saw improved TIR, PPG, no increase in hypoglycemia, or weight gain. There was less glucose variability and time in hypoglycemia.
  • Education is key about shorter duration of action as patients become non-adherent due to high doses.
  • This was a very small study of 60 people; larger studies with more power are needed.


Akturk, Halis Kaan, et al. “Improved Postprandial Glucose with Inhaled Technosphere Insulin Compared with Insulin Aspart in Patients with Type 1 Diabetes on Multiple Daily Injections: The STAT Study.” Diabetes Technology & Therapeutics, 2018, doi:10.1089/dia.2018.0200.

Arsalan Hashmi, PharmD. Candidate, LECOM College of Pharmacy