A Randomized Comparison of Postprandial Glucose Excursion Using Inhaled Insulin Versus Rapid-Acting Analog Insulin in Adults With Type 1 Diabetes Using Multiple Daily Injections of Insulin or Automated Insulin Delivery Free

In individuals with type 1 diabetes, postmeal hyperglycemia is difficult to prevent with subcutaneous rapid-acting analog (RAA) insulin, even with automated insulin delivery (AID) systems. Technosphere insulin inhalation powder (TI) (Afrezza; MannKind Corporation, Danbury, CT) has a more rapid onset of action, and serum insulin levels return to baseline levels sooner than with RAA insulin (1,2). At the start of a multicenter 17-week randomized trial comparing a regimen of TI plus insulin degludec versus usual care in adults with type 1 diabetes, we performed a standardized in-clinic meal challenge to compare the effects of TI versus RAA insulin on postmeal glycemia.

The study was conducted at 19 sites in the U.S. (7 community-based sites and 12 academic centers). The protocol and Health Insurance Portability and Accountability Act–compliant informed consent forms were approved by a central institutional review board, and informed consent was obtained from each participant. The study is registered at ClinicalTrials.gov (NCT05904743).

Study participants were ≥18 years old with type 1 diabetes and used AID, a nonautomated pump, or multiple daily injections (MDI) plus continuous glucose monitoring (CGM). Additional inclusion and exclusion criteria are listed in Supplementary Table 1.

Participants were fasting with no manual insulin boluses for 4 h and were expected to be wearing a blinded Dexcom G6 Pro sensor (Dexcom, Inc.) during the meal challenge to provide glucose data for the main analyses. If CGM data were not available, blood glucose (BG) measurements were used for analyses. Participants were randomly assigned (1:1) to receive either TI or RAA. The TI dose during the meal challenge was the first dose that participants received. Pump use was continued, including AID insulin delivery. In the TI group, AID users did not enter carbohydrates into the pump, and to safely avoid insulin dose accumulation, Tandem Control-IQ users placed the pump in sleep mode (to avoid autoboluses) for at least 4 h postmeal.

The glucose level was expected to be 91–219 mg/dL 15 min prior to the consumption of a BOOST nutritional shake (approximately 240 calories, 37 g carbohydrate, 10 g protein, 4 g fat). In the RAA group, the dose of insulin was based on the participant’s usual bolus for that amount of carbohydrate. In the TI group, the dose of inhaled insulin was determined by rounding the usual RAA dose to the nearest whole number, multiplying by 2, and then rounding down to the nearest 4-unit dose of TI (not to exceed 24 units). Dose determinations were made without glucose level correction. RAA was given 10 ± 5 min before the start of the meal and TI at the start of the meal.

BG measurements were obtained, with a McKesson True Metrix Pro professional monitoring blood glucose meter, 15 min before the meal was ingested, at the start of the meal, and 15, 30, 45, 60, 90, and 120 min after the start of the meal. During or at the completion of 2 h postmeal, for a BG value <70 mg/dL, 15 g oral carbohydrate was ingested; for a BG measurement >400 at any time or >300 mg/dL for 1 h, additional insulin was administered (RAA insulin for the RAA group and TI for the TI group). After the first 2 h, the participant was discharged from the clinic and asked not to have a meal or snack until 4 h after the start of the meal (unless needed to treat hypoglycemia) and to continue to wear the blinded sensor.

Glucose data from the 2 h after the start of the meal were used in the primary analysis. Additional analysis evaluated the glucose data for the 4 h after the meal challenge and in subgroups based on prestudy insulin delivery method. The primary outcome was the 2-h area under the curve >180 mg/dL (AUC180), which represents the area from the glucose values to the 180-mg/dL threshold, with glucose values ≤180 mg/dL having an area of 0. The area was divided by the time frame duration to yield standardized results. This end point was selected to be primary as it accounts not only for time glucose levels are >180 mg/dL but also the magnitude of the excursion >180 mg/dL. Additional key outcomes included maximum glucose excursion from baseline, peak glucose, and time to peak glucose. Mean glucose and CGM metrics for percent time in the standard ranges also were computed. Treatment group comparisons were made using a linear mixed-effects regression model adjusting for the time 0 glucose value, baseline HbA_1c, insulin delivery method, and clinical site. To convert glucose values from milligrams per deciliter to millimoles per liter, multiply by 0.05551. SAS version 9.4 was used for the analyses.

The meal challenge was completed by 122 participants (61 in each group). Participant characteristics appeared balanced between groups (Table 1). An AID system was being used by 48%, a predictive low glucose suspend pump by 2%, a nonautomated pump by 6%, and MDI by 43%. Supplementary Tables 2 and 3 provide information about the meal challenges and insulin doses that were given.

Mean 2-h AUC180 was 40 ± 44 mg/dL in the TI group versus 50 ± 42 mg/dL in the RAA group (adjusted difference −12 mg/dL, 95% CI −22 to −2, P = 0.02). With TI, the glucose excursion was smaller (P = 0.01), peak glucose was lower (P = 0.01), and time to peak glucose was shorter (P = 0.006) (Table 2 and Fig. 1A). This glucose excursion pattern was observed when starting glucose was <150 mg/dL vs. ≥150 mg/dL, although it appeared larger with the latter (Supplementary Fig. 1). CGM metrics reflecting hyperglycemia trended similarly, favoring TI (Supplementary Table 4). The treatment group difference gradually dissipated after peaking at about 90 min, and by 3 h the glucose levels in the two groups appeared similar (Fig. 2 and Supplementary Tables 5 and 6). A reduction in postmeal hyperglycemia with TI compared with RAA was suggested for both non-AID and AID users, although the effect appeared to be greater in non-AID users (Fig. 1B and C and Table 2). Results during the first 2 h appeared similar when just BG values were used in the analysis (Supplementary Fig. 2).

During the 2 h of the in-clinic monitored procedure, a BG value <70 mg/dL occurred in one participant in each group. Additional insulin was given for prolonged hyperglycemia due to an infusion set failure in one RAA participant. One participant in the TI group had a second TI dose after 90 min due to investigator discretion. One participant developed coughing, neck tingling, throat tightness, and breathing difficulties after TI inhalation that dissipated without intervention.

During the 4-h postmeal period, mean CGM glucose levels <70 mg/dL were low with both TI and RAA (0.0% ± 0.0% and 0.2% ± 1.7%, respectively). CGM glucose <54 mg/dL occurred in one (2.3%) participant in the TI group and two (4.5%) participants in the RAA group (Supplementary Table 6).

In this study of adults with type 1 diabetes predominately using AID or MDI, postmeal hyperglycemia was significantly less with TI than with RAA insulin. Given the known pharmacokinetics of TI and RAA, the results are not surprising (1). Trends favoring TI were observed in both AID and non-AID users, although trends for non-AID users were more favorable. The results observed in AID users are consistent with a prior study of AID users that compared TI with RAA (3).

The current labeling of TI bases initial dosing on an approximate 1:1 conversion of RAA dose. However, the current study and a prior study (4) as well as clinical use of TI have indicated that a conversion of at least 2:1 represents a bioequivalent TI dose relative to RAA that can be inhaled safely with substantially greater reduction in postmeal hyperglycemia. In the current study, postmeal hypoglycemia was similarly infrequent in both groups.

One benefit of TI compared with RAA is that it can be dosed right at the start of the meal due to its rapidity of action rather than 5–15 min prior to the meal for optimal RAA dosing. Additionally, the more rapid dissipation of effect of TI compared with RAA allows for repeat bolusing after 1 h if the glucose level is elevated without concern for insulin stacking, as would be the case with RAA redosing. Since repeat dosing after 1 h was not done during the meal challenge, the benefit of TI observed during the first 2 h was, as expected, not sustained through 4 h.

Strengths of the study include a multicenter randomized trial design, inclusion of both MDI and AID system users, and a standardized meal protocol. RAA insulin was given 5–15 min prior to the meal, whereas TI was inhaled at the start of the meal to mirror clinical recommendations. Limitations include the lack of study CGM data during the meal challenge for some participants. However, sensitivity analyses limited only to participants with CGM data and analyses using only BG data produced similar results, suggesting that combining CGM and BG data did not affect the results.

In summary, in a cohort of adults with type 1 diabetes that included approximately 50% AID users, postmeal hyperglycemia was less with TI than with RAA insulin.

Acknowledgments. J.B.B. is an editor of Diabetes Care but was not involved in any of the decisions regarding review of this manuscript or its acceptance.

Funding. INHALE-3 (Afrezza Combined With Insulin Degludec Versus Usual Care in Adults With Type 1 Diabetes) was supported by funding from MannKind to the JAEB Center for Health Research Foundation, Inc., which then provided funding to the clinical sites.

The content is solely the responsibility of the authors. MannKind reviewed the manuscript and provided comments but had no rights of approval of the content of the manuscript.

Duality of Interest. I.B.H. reports consulting for Abbott, Roche, Hagar, and Vertex Pharmaceuticals and research support from MannKind, Dexcom, and Tandem Diabetes Care. R.W.B. reports no personal financial disclosures but reports that his institution has received funding on his behalf as follows: grant funding, study supplies, and consulting fees from Insulet, Tandem Diabetes Care, and Beta Bionics; grant funding and study supplies from Dexcom; grant funding from Bigfoot Biomedical; study supplies from Medtronic, Ascencia, and Roche; consulting fees and study supplies from Eli Lilly and Novo Nordisk; and consulting fees from embecta, Sequel Med Tech, Vertex Pharmaceuticals, Hagar, Ypsomed, Sanofi, and Zucara. A.D.C. reports consulting for Novo Nordisk, Intuitive Surgical, Inc., Hanmi-IQVIA, and Wellworksforyou, Inc. A.B. reports ongoing research grant support from the following: Boehringer Ingelheim Pharmaceuticals, Inc., Abbot Diabetes Care, AbbVie, Inc., Covance, Inc., Dexcom, Inc., Eli Lilly and Company, Madrigal Pharmaceuticals, Inc., Medtronic, Novo Nordisk, Carmot, Quintiles, Akero, 89 Bio, Inc., Viking Therapeutics, Barbara Davis Center for Diabetes, PPD, Zydus, Kowa Pharmaceuticals America, Inc., until October 2023, and Insulet Corporation until June 2022. A.T. receives supplies and research support provided to her institution from the following: Tandem Diabetes Care, Altimmune, Insulet, Dexcom, Medtronic, Eli Lilly, MannKind, Roche, and Abbott Diabetes Care. C.L. receives research support from Tandem Diabetes Care and Dexcom paid to her institution. C.J.L. reports research support paid to her institution from Dexcom, Tandem Diabetes Care Insulet, Novo Nordisk, and Abbott Diabetes and consulting fees from Dexcom and Tandem Diabetes Care. C.J. reports consulting for Medtronic advisory boards and speaking for Medtronic on insulin pumps. D.P. reports support from Novo Nordisk, Eli Lilly, MannKind, AstraZeneca, and Boehringer Ingelheim. D.W.S. reports consulting for Abbott Diabetes Care and research funding through his institution from the following: Abbott Diabetes Care, Tandem Diabetes Care, Novo Nordisk, and the National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases. G.M. reports research support from Dexcom, Abbot, Tandem Diabetes Care, Insulet, and Novo Nordisk. H.K.A. reports grant support received for his institution from Medtronic, Dexcom, and Tandem Diabetes Care and reports honorarium consulting from Medtronic, Dexcom, and Tandem Diabetes Care. J.B.B. reports grant support from Bayer, Boehringer-Ingelheim, Carmot, Corcept, Dexcom, Eli Lilly, Insulet, MannKind, Novo Nordisk, and vTv Therapeutics; consulting contracts from Alkahest, Altimmune, Anji, Aqua Medical, Inc., AstraZeneca, Boehringer-Ingelheim, CeQur, Corcept Therapeutics, Dasman Diabetes Care (Kuwait), Eli Lilly, embecta, Fortress Biotech, GentiBio, Glyscend, Insulet, Mediflix, Medscape, Mellitus Health, Metsera, Moderna, Novo Nordisk, Pendulum Therapeutics, Praetego, ReachMD, Stability Health, Tandem Diabetes Care, Terns, Inc., and Vertex Pharmaceuticals; expert witness engagement by Medtronic MiniMed; and stock options from Glyscend, Mellitus Health, Pendulum Therapeutics, Praetego, and Stability Health. K.R.K. has received consulting fees from Novo Nordisk. K.Ca. receives research support provided to her institution from Dexcom, Abbott, Medtronic, Eli Lilly, MannKind, and Insulet and receives consulting fees from Dexcom. M.K. reports acting on the clinical events committee for Tandem Diabetes Care; acting on the advisory board for Quest Diagnostics and Corcept; and research support from 89 Bio, Diamyd, Abbott, Akero, Kowa, Zydus, Biolinq, Corcept, Dexcom, Eli Lilly, Gilead, Insulet, Ionis, MannKind, Novo Nordisk, Pfizer, Reata, Inventiva, and Tandem Diabetes Care. O.H. reports consultant fees for Abbot Nutrition and receives research grants for his institution from Novo Nordisk, Eli Lilly, and Cambridge Technology. R.S.W. has participated in clinical trials, through her institution, sponsored by Tandem Diabetes Care, Insulet, Eli Lilly, Amgen, and Diasome and has received devices from Dexcom for clinical trials through her institution at a reduced cost. T.B. reports acting on a speaker bureau for Eli Lilly, Novo Nordisk, Sanofi, Boehringer-Ingelheim, AstraZeneca, and Dexcom; acting on the advisory board for Insulet and MannKind; and receiving research support form Eli Lilly, Novo Nordisk, Dexcom, Abbott, Medtronic, Tandem Diabetes Care, Biomea, and MannKind. Y.C.K. reports product support from Dexcom. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. All authors contributed to the interpretation of the study results and provided critical review of the content of the manuscript. R.W.B. and I.B.H. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. The results were presented at the 17th International Conference of the Advanced Technologies & Treatments for Diabetes (ATTD), 6–9 March 2024, Florence, Italy.

Handling Editors. The journal editor responsible for overseeing the review of the manuscript was Mark A. Atkinson.