Efficacy and Safety of Inhaled Technosphere Insulin (TI) vs. Rapid-Acting Analog in Youth with HbA1c ≤ 9.5%: Subgroup Analysis from INHALE-1

Authors: Kevin Kaiserman, et al.

This subgroup analysis of the INHALE-1 study included 209 children and adolescents with type 1 or type 2 diabetes and baseline glycated hemoglobin (HbA1c) ≤9.5% who received either inhaled Technosphere Insulin (TI; n=106) or rapid-acting insulin analogs (RAA; n=103), in combination with basal insulin, for 26 weeks. At Week 26, there was a change in the HbA1c in both groups. The least squares mean change from baseline was 0.19% (95% confidence interval [CI] 0.007–0.365) with TI and 0.08% (95% CI -0.103 to 0.256) with RAA. The between-group difference was 0.11% (95% CI -0.145 to 0.364), meeting the predefined criterion for non-inferiority (CI upper limit <0.4%) (CI upper limit <0.4%), indicating that glycemic control with TI was comparable to that achieved with RAA in this subgroup. Safety outcomes were similar between treatments. Pulmonary function remained stable over the 26-week study period, and continuous glucose monitoring metrics, including hypoglycemia, showed no meaningful differences between groups. In the overall study population (N=230), the non-inferiority endpoint was not met. However, a sensitivity analysis excluding one participant with a protocol deviation met the non-inferiority criteria.

Overall, inhaled TI demonstrated efficacy and safety comparable to RAA in youth with diabetes and baseline HbA1c ≤9.5%, supporting its use as an alternative prandial insulin option.

These findings highlight the importance of aligning treatment strategies with patients’ preferences, readiness, and goals, as greater engagement and comfort with therapy may support improved clinical outcomes.

Efficacy and Safety of Inhaled vs. Subcutaneous Insulin: A Meta-Analysis

Authors: Ali Raza Khan, et al

This meta-analysis evaluated the efficacy and safety of inhaled insulin compared with subcutaneous (SQ) insulin in patients with type 1 and type 2 diabetes mellitus. A total of 19 randomized controlled trials (RCTs) were included in the analysis. Compared with SQ insulin, inhaled insulin was associated with a greater reduction in fasting blood glucose levels (mean difference [MD] -22.12; 95% confidence interval [CI] -29.29 to -14.95). The proportion of patients achieving glycated hemoglobin (HbA1c) <7% was similar between treatment groups, indicating comparable overall glycemic control. Patients receiving inhaled insulin also experienced significant weight reduction compared with those receiving SQ insulin. However, inhaled insulin was associated with small but significant declines in pulmonary function measures, including diffusing capacity of the lungs for carbon monoxide (DLCO) (MD -0.44; 95% CI -0.78 to -0.11) and forced expiratory volume in one second (FEV1) (MD -0.03; 95% CI -0.04 to -0.01).

Overall, inhaled insulin provided glycemic control comparable to SQ insulin, while offering greater reductions in fasting blood glucose and body weight. However, additional research is needed to assess its impact on pulmonary function and strengthen its safety profile.

Exploratory Evaluation of Technosphere Insulin with Automated Insulin Delivery: Impact of Total Daily Dose Algorithms

Authors: Joanne K Rinker, et al.

This exploratory analysis evaluated glycemic outcomes in adults with type 1 diabetes who used inhaled Technosphere Insulin (TI) for mealtime and correction dosing alongside automated insulin delivery (AID) systems providing basal insulin. The study assessed whether outcomes differed depending on whether the AID algorithm was based on total daily insulin dose (TDD).

Nine participants were included, of whom six used TDD-dependent AID algorithms and three used algorithms that were not dependent on TDD. Glycated hemoglobin (HbA1c) changes were assessed after 90 days of treatment. HbA1c improved in both groups during the study period. Participants using TI with AID systems that were not dependent on TDD showed a slightly greater reduction in HbA1c compared with those using TDD-dependent systems. Mean HbA1c decreased by 0.33% in the non-TDD-dependent group and by 0.22% in the TDD-dependent group after 90 days.

The findings suggest that glycemic outcomes with TI may vary depending on the type of AID algorithm used. However, the small sample size and exploratory nature of the study limit interpretation, and larger studies are needed to better understand the interaction between TI and different AID systems.

Inhaled Insulin Assessment of Incident Lung Cancer Risk among Adults with Type 2 Diabetes (T2D) Compared with Other Therapies for Diabetes Using Real-World Evidence

Authors: Kevin Kaiserman, et al.

This retrospective real-world evidence study evaluated the occurrence of lung cancer among adults with type 2 diabetes (T2D) treated with Technosphere Insulin (TI), the only inhaled insulin currently approved by the US Food and Drug Administration (FDA). The analysis used linked health insurance claims and electronic medical record data collected between January 2017 and August 2025. The study included 7,494,072 adults with T2D across the United States, of whom 2,056 received TI. The primary comparison was between TI users and patients receiving injectable rapid-acting insulin (RAA). Additional comparator groups included patients receiving no treatment, oral glucose-lowering agents, glucagon-like peptide-1 receptor agonists (GLP-1s), non-insulin therapies, and long-acting insulin. Among TI users, 10 cases of incident lung cancer were identified, corresponding to an incidence proportion of 0.49%. In comparison, lung cancer incidence was 1.17% among 8,936 patients receiving RAA. Incidence proportions in the other cohorts were 1.12% for no treatment, 0.81% for oral agents, 0.68% for non-insulin therapies, 0.78% for long-acting insulin, and 0.33% for GLP-1 therapies. Patients receiving TI had a mean follow-up duration of 1,941.9 days, which was longer than that reported for any comparator cohort.

Overall, this interim analysis did not identify an increased occurrence of lung cancer among TI users compared with other treatment groups in adults with T2D.

Inhaled Insulin Demonstrates Earlier Completion of Total Pharmacodynamic Effect Compared with Lispro

Authors: Kevin Kaiserman, et al.

This analysis examined the timing of insulin action for inhaled Technosphere Insulin (TI) and subcutaneous insulin lispro in adults with type 1 diabetes. Rather than evaluating overall pharmacodynamic activity alone, the study assessed how quickly each insulin achieved its total glucose-lowering effect over time. The comparison used TI 12 U and insulin lispro 8 U, as these doses provided similar overall pharmacodynamic exposure The cumulative percentage of total glucose-lowering effect achieved at different time points was calculated using glucose infusion rate data from a glucose clamp study. TI demonstrated a substantially earlier onset of action than insulin lispro. By 60 minutes, TI had delivered 53% of its total pharmacodynamic effect compared with 10% for insulin lispro, a difference of 43 percentage points. At 90 minutes, TI had achieved 75% of its total effect versus 27% with lispro, while at 120 minutes the corresponding values were 88% and 45%, respectively. The proportion of remaining insulin effect was consistently lower with TI. At 60 minutes, 47% of TI activity remained compared with 90% for lispro; at 90 minutes, 25% versus 73%; and at 120 minutes, 12% versus 55%, respectively.

These findings indicate that TI delivers most of its glucose-lowering effect within the first 1–2 hours after dosing, whereas insulin lispro exerts a larger proportion of its effect later in the postprandial period, indicating potential for earlier dosing adjustments with a reduced risk of insulin stacking and subsequent hypoglycemia.

Inhaled Insulin Demonstrates Lower Variability and Faster Onset Compared with Subcutaneous Rapid-Acting Analogs

Authors: Jennifer Nguyen, et al.

This analysis from the INHALE-3 study examined whether higher individualized dosing of inhaled Technosphere Insulin (TI) was associated with improved postprandial glucose control in adults with type 1 diabetes (T1D). Participants receiving TI completed a standardized meal challenge before and after 17 weeks of dose titration, while those receiving usual care (UC) completed the same challenge using their rapid-acting insulin analog regimen. Based on insulin dose per kilogram body weight (U/kg), participants were categorized into low- and high-dose groups. Mean doses were 0.074 U/kg (TI-low) and 0.16 U/kg (TI-high) for TI, and 0.042 U/kg (UC-low) and 0.084 U/kg (UC-high) for usual care. Participants in the TI-high group experienced significantly lower postprandial glucose excursions than those in the UC-high group from 30 to 120 minutes after the meal (p<0.05). Glucose excursions were also significantly lower in the TI-high group compared with the TI-low group (p<0.05). No significant differences were observed between the UC-low and UC-high groups or between the TI-low and UC-low groups. Overall, these findings underscore the importance of individualized TI dosing in optimizing postprandial glycemic control in type 1 diabetes. While lower doses may achieve outcomes comparable to standard care, higher, appropriately titrated doses appear necessary for improved postprandial coverage in some individuals. This highlights the potential of a tailored dosing approach to better match real‑world insulin needs and reinforces the need for further research to refine patient‑specific TI titration strategies.

Inhaled Technosphere Insulin (TI) Compared with Rapid-Acting Analog Insulin (RAA) in Gestational Diabetes (GDM)

Authors: Amy Valent, et al.

This interim analysis evaluated the effects of inhaled Technosphere Insulin (TI) and rapid-acting insulin analogs (RAA; lispro or aspart) on postprandial glucose levels in pregnant women with gestational diabetes mellitus (GDM). Pregnant women with GDM using CGM took part in two identical breakfast sessions (~45 g carbs), receiving either RAA or TI (about twice the RAA dose) within 10 days. Glucose levels were checked every 15–30 minutes for at least 3 hours, and this analysis compares results from the first 10 participants.

Participants had a mean age of 33±6 years, a pre-pregnancy body mass index (BMI) of 35±6 kg/m², and a mean gestational age of 29±4 weeks. Half of the participants identified as Hispanic or Latinx. Following the standardized breakfast meal, TI showed a trend toward lower postprandial glucose excursions compared with RAA. Mean glucose excursion was 2.8 mg/dL lower with TI. The area under the glucose curve above 120 mg/dL and above 140 mg/dL was also lower with TI by 2.7 mg/dL and 1.3 mg/dL, respectively. One-hour postprandial glucose levels were lower with TI than with RAA (121±16 mg/dL vs 129±25 mg/dL), while 2-hour postprandial glucose levels were similar between groups (108±20 mg/dL vs 107±25 mg/dL). No hypoglycemic events occurred during the TI meal sessions. In contrast, hypoglycemia occurred during 3 meal sessions in the RAA group.

Overall, TI provided postprandial glucose control comparable to RAA, with a trend toward lower glucose excursions and fewer hypoglycemic events in this interim analysis.

Participants Achieving HbA1c <8% in Youth Report Greater Treatment Satisfaction with Inhaled Technosphere Insulin vs. Rapid-Acting Analogs

Authors: Joanne K Rinker, et al.

This post hoc analysis of the INHALE-1 study examined treatment satisfaction among children and adolescents with type 1 or type 2 diabetes who achieved glycated hemoglobin (HbA1c) levels below 8% after 26 weeks of treatment. Patient-reported outcomes were assessed using the Diabetes Treatment Satisfaction Questionnaires (DTSQs) completed by adolescents and by parents of younger children. Previous results from the study showed similar HbA1c outcomes with inhaled Technosphere Insulin (TI) and multiple daily injections of rapid-acting insulin analogs (RAA). However, among participants who achieved HbA1c <8%, treatment satisfaction improved significantly more with TI than with RAA. On the Teen DTSQs, participants receiving TI (n=28) reported greater improvement in treatment satisfaction compared with those receiving RAA (n=26; p=0.02). Similarly, parent-reported satisfaction scores improved more in the TI group (n=13) than in the RAA group (n=23; p=0.01).

These findings suggest that although glycemic outcomes were similar between treatments, youth who achieved glycemic targets with TI reported higher treatment satisfaction than those using injectable RAA therapy.

ADA 2026, June 5-8, New Orleans