Pharmacological characteristics of once‐weekly insulin icodec in Japanese individuals with type 1 diabetes

ABSTRACT

Introduction

Insulin icodec is a basal insulin designed for once‐weekly administration. This study assessed the pharmacological properties of icodec in Japanese individuals with type 1 diabetes (T1D).

Materials and Methods

In a randomized, open‐label, crossover study, 24 Japanese individuals with T1D (20–64 years; glycated hemoglobin ≤9.0%) received once‐weekly icodec for 8 weeks and once‐daily insulin glargine U100 for 14 days at individual constant equimolar doses per week together with bolus insulin aspart. Individual doses were determined during run‐in with glargine U100 titrated to prebreakfast self‐measured plasma glucose (SMPG) of 4.4–7.2 mmol/L. Blood samples for icodec pharmacokinetics were taken from the first icodec dose until 35 days after last dose. The steady‐state glucose‐lowering effect was measured in glucose clamps (target 6.7 mmol/L) during 24–48 h and 150–168 h after last icodec dose and 0–24 h after last glargine U100 dose. One‐week glucose‐lowering effect of icodec was simulated using a pharmacokinetic/pharmacodynamic model. Hypoglycemia was identified from SMPG during the treatment periods.

Results

Icodec pharmacokinetic steady state was achieved on average after 2–3 weeks of treatment. Model‐derived daily glucose‐lowering effect during the weekly dosing interval averaged 14.6%, 18.0%, 16.6%, 14.9%, 13.3%, 11.9%, and 10.7%, respectively. Rates of level 2 hypoglycemia (PG <3.0 mmol/L) were 37.3 vs 30.6 episodes per patient‐year of exposure for icodec vs glargine U100.

Discussion

Icodec pharmacological properties in Japanese individuals with T1D in this study support the potential of icodec to provide basal insulin coverage with once‐weekly dosing in Japanese individuals with diabetes.

This study assessed the pharmacokinetic and pharmacodynamic properties of once‐weekly basal insulin icodec in Japanese individuals with type 1 diabetes. The results of the study support the potential of insulin icodec to provide basal insulin coverage with once‐weekly dosing in Japanese individuals with diabetes.

INTRODUCTION

Basal insulin products for managing diabetes have until recently been developed for once‐ or twice‐daily administration. In individuals with type 2 diabetes (T2D), basal insulin initiation may be delayed due to treatment barriers such as the risk of hypoglycemia and weight gain, fear of injections and the treatment burden associated with daily administration ¹ , ² . In type 1 diabetes (T1D), nonadherence to basal insulin treatment has been shown to adversely affect glycemic control, potentially increasing the occurrence of microvascular complications, and to raise the risk of diabetic ketoacidosis ³ , ⁴ , ⁵ . Once‐weekly basal insulin reduces the required basal insulin injections from at least 365 to 52 per year. This may reduce therapeutic inertia, increase treatment adherence, improve the effect on glycemic control, and potentially decrease the risk of diabetic ketoacidosis ⁶ , ⁷ , ⁸ .

Insulin icodec is a novel basal insulin with pharmacokinetic and pharmacodynamic properties suitable for once‐weekly administration, as shown in White individuals, including a half‐life of approximately 1 week ⁹ , ¹⁰ , ¹¹ . After subcutaneous administration, icodec is absorbed into the bloodstream, where it binds to albumin creating an essentially inactive depot of icodec ⁹ , ¹² . Slowly over time, icodec can activate the insulin receptor, however with low affinity, which further contributes to its slow clearance and steady glucose‐lowering effect ⁹ . In Phase 3a trials in individuals with T2D, icodec showed superiority in glycated hemoglobin (HbA_1c) reduction vs once‐daily basal insulin in a basal‐only regimen with low rates of clinically significant or severe hypoglycemia for all treatment arms, and noninferiority in change in HbA_1c vs once‐daily insulin glargine U100 in a basal‐bolus regimen with similar rate of clinically significant or severe hypoglycemia between treatment arms ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ . During the 26‐week main part of the Phase 3a trial in T1D, icodec showed noninferiority in change in HbA_1c vs once‐daily insulin degludec, however with higher rate of clinically significant or severe hypoglycemia for icodec vs degludec (19.9 vs 10.4 episodes per patient‐year of exposure [PYE]) ¹⁸ . Still, rates were lower than those previously reported in trials with degludec in T1D (42.5–88.3 episodes per PYE) ¹⁹ , ²⁰ , ²¹ , ²² .

The pharmacokinetic and/or pharmacodynamic properties of drugs including insulin may be affected by race and ethnicity ²³ , ²⁴ , ²⁵ . The aim of this study was for the first time to assess the pharmacological characteristics of icodec in Japanese individuals and relate the results to those in White individuals. Although icodec can be used as basal insulin in both T1D and T2D, this study enrolled participants with T1D only, in order to reduce potential interference from endogenous insulin on the assessment of pharmacodynamics.

MATERIALS AND METHODS

Study design

This randomized, single‐center (Hakata Clinic, Fukuoka, Japan), open‐label, two‐period crossover study (ClinicalTrials.gov identifier: NCT03766854) was conducted according to the Declaration of Helsinki ²⁶ and ICH Good Clinical Practice ²⁷ . The protocol for this study was approved by an appropriate institutional review board (Hakata Clinic IRB, Fukuoka, Japan).

Participants

Participants were Japanese between the ages of 20 and 64 years, diagnosed with T1D for at least 1 year, and either receiving multiple daily insulin administrations or continuous subcutaneous insulin infusion with a daily basal insulin dose of at least 0.2 U/kg/day, with an HbA_1c level of maximum 9.0% (75 mmol/mol), a fasting C‐peptide level below 0.3 nmol/L, and BMI between 18.5 and 28.0 kg/m². People were excluded from participation if they had a history or presence of any clinically significant health issues related to the respiratory, metabolic, renal, hepatic, gastrointestinal, or endocrine systems (excluding those linked to diabetes), blood pressure readings outside 90–139 mmHg for systolic or 50–89 mmHg for diastolic, frequent severe hypoglycemia (more than one episode in the last 180 days), hypoglycemia unawareness, an estimated glomerular filtration rate below 60 mL/min/1.73 m², an alanine aminotransferase level of at least 2.5 times the upper normal limit, a bilirubin level more than 1.5 times the upper normal limit, were pregnant or currently treated with systemic corticosteroids, monoamine oxidase inhibitors, systemic nonselective beta‐blockers, or growth hormone. Before engaging in any study‐related activities, written informed consent was provided by all participants.

Procedures and assessments

Following screening, the study comprised a run‐in period, two treatment periods and a follow‐up (Figure S1).

During run‐in (ranging from 2 to 70 days), individuals transitioned from their pre‐study insulin therapy to a daily insulin glargine U100 administration at 20:00 h, along with insulin aspart as bolus insulin. The glargine U100 dose was titrated individually to achieve a self‐measured plasma glucose (SMPG) of 4.4–7.2 mmol/L before breakfast. Once the average of three daily consecutive SMPG values before breakfast fell within this range, participants proceeded to randomization and the first treatment period.

During the treatment periods, participants received once‐weekly icodec for 8 weeks and once‐daily glargine U100 for 14 days, with the order of treatment being randomized. The difference in treatment duration for icodec vs glargine U100 was due to their different dosing frequency and time to steady state and the need for pharmacokinetic and pharmacodynamic evaluations at steady state. Both icodec and glargine U100 were given at a consistent dose, with dose adjustments only being allowed for safety reasons. Icodec (4,200 nmol/mL; Novo Nordisk, Bagsværd, Denmark) was dosed subcutaneously in the left thigh around 20:00 h on the same day of the week by trained staff using a PDS290 prefilled pen injector (Novo Nordisk). The weekly dose of icodec was calculated as seven times the individual's daily basal dose of glargine U100 determined during the run‐in period. To account for the time required to reach icodec steady state, participants were given additional basal insulin (glargine U100) during the first 2 weeks of icodec treatment. The amount of glargine U100 given during this time was adjusted by the investigator based on the participant's SMPG levels before breakfast. Glargine U100 (100 U/mL; Sanofi, Paris, France) was self‐administered subcutaneously in the right thigh around 20:00 h by the participant using a SoloStar pen injector (Sanofi), except that the final dose before the glucose clamp was administered by the study staff. The daily dose of glargine U100 was set to the dose determined during the run‐in period. Buffer periods were introduced between the two treatment periods. These were between 35 and 49 days following icodec treatment and between 2 and 15 days following glargine U100 treatment. During these buffer periods, participants were given once‐daily glargine U100 as their basal insulin at a dose similar to that established during the run‐in period, albeit with a reduced dose during the first 2 weeks after icodec treatment as adjusted by the investigator based on the participant's SMPG levels before breakfast (approximately 75% reduction in the first week and 50% reduction in the second week).

The follow‐up visit took place between 39 and 45 days following the final icodec administration, or between 5 and 11 days following the final glargine U100 administration.

The pharmacokinetic characteristics of icodec were assessed through blood sampling at prespecified time points from start of icodec treatment up to 35 days following the final icodec administration (Table S1). Total serum icodec levels (i.e., icodec bound to albumin plus unbound icodec) were determined by a validated icodec‐specific immunoassay ²⁸ .

Icodec and glargine U100 pharmacodynamics were evaluated at steady state using glucose clamps (STG‐55; Artificial Pancreas, NIKKISO Co. Ltd., Japan) following the final dose in each treatment period. For icodec, two glucose clamps were performed between 24 and 48 h and between 150 and 168 h following the final icodec dose. These intervals spanned the anticipated time of maximum glucose infusion rate and the latter part of the one‐week dosing interval. For glargine U100, a single glucose clamp was carried out over the entire 24‐h dosing interval. Before each glucose clamp, participants were adviced to refrain from using over‐the‐counter medications for the previous 2 weeks. The use of routine vitamins and occasional salicylates was permissible up to the last 48 h. Additionally, participants were to avoid smoking, intense physical activity and intake of alcohol, coffee, tea, chocolate, and beverages containing methylxanthine for the last 48 h. It was also important to prevent hypoglycemia during the last 24 h. Lastly, participants had to fast and avoid using bolus insulin during the last 8 h. During a 5‐h clamp run‐in period, participants were given an adjustable intravenous insulin aspart infusion (100 U/mL; Novo Nordisk; 40 U in 98 mL saline and 2 mL of the participant's blood) or glucose (10%) to achieve a target blood glucose level of 6.7 mmol/L. The aim was for blood glucose to stay within a range of 6.7 mmol/L ± 20% between 60 and 30 min prior to clamp start and within 6.7 mmol/L ± 10% in the last 30 min before clamp start. Any deviations from these targets were allowed for no more than 5 min during every 30‐min period. Moreover, there should be no more than a 0.3 mmol/L fluctuation in blood glucose levels in the last 10 min before clamp start. If insulin was infused, it was reduced as much as possible and stopped before clamp start. Throughout the clamp, participants were kept fasting, except for water when necessary, and were positioned either supine or semi‐supine. The clamp was stopped sooner than scheduled if blood glucose consistently rose above 11.1 mmol/L without the need for glucose infusion for the last 30 min.

Safety evaluations comprised adverse events (AEs), hypoglycemic episodes categorized as Level 2 (clinically significant; plasma glucose <3.0 mmol/L) and Level 3 (severe cognitive impairment requiring external assistance for recovery) ²⁹ , hyperglycemic episodes (plasma glucose >16.7 mmol/L), vital signs, electrocardiogram, physical examination, and clinical laboratory assessments. Hypoglycemic and hyperglycemic episodes were identified from SMPG measurements. Any clinically significant results in vital signs, electrocardiogram, physical examination, and clinical laboratory assessments were recorded as AEs.

Statistical analysis

The primary end point was total icodec exposure during the one‐week dosing interval at steady state (AUC_τ,SS). AUC_τ,SS was used in a post hoc evaluation of dose‐proportionality. Assuming a standard deviation (SD) for log(AUC_τ,SS) of 0.25, derived from a prior multiple‐dose study with icodec in White people ⁹ , 20 participants implied that the 95% confidence interval (CI) for the geometric mean AUC_τ,SS would have a width of within 28.2% of the geometric mean with 90% probability. This level of precision was deemed adequate for the study's aims. To ensure that at least 20 participants could be evaluated, it was scheduled that 24 participants were randomized.

The mean measured serum icodec concentration–time curve during a one‐week dosing interval at steady state was derived after the final icodec administration in the treatment period. Pharmacokinetic end points for icodec comprised AUC_τ,SS, which was calculated by integrating the serum icodec concentration–time profile using a linear trapezoidal method from 0 to 168 h after dosing, the observed maximum concentration at steady state (C _max,SS), and the terminal half‐life at steady state (t _½,SS), which was derived as log(2) divided by the estimated terminal rate constant using linear regression (λ _z,SS). Evaluation of icodec dose‐proportionality with respect to AUC_τ,SS was carried out as described previously ¹¹ .

Data on the individual glucose infusion rates for icodec and glargine U100 at steady state were smoothed using the Loess smoothing method using a smoothing factor of 0.25.

Safety evaluations were summarized using descriptive statistics.

Pharmacokinetic‐pharmacodynamic modeling

The glucose‐lowering effect of icodec over a full one‐week dosing interval at steady state was estimated for every participant by pharmacokinetic‐pharmacodynamic modeling as previously outlined ¹¹ . The only modifications were that the pharmacodynamic model did not include a threshold parameter and that inter‐individual variability was not included on the turnover parameter.

RESULTS

Participant disposition, demographics, and exposure

A total of 29 individuals went through screening and 24 of these were randomized, exposed to icodec and glargine U100 and completed the study (Figure S2). Demographics are provided in Table S2. During the treatment periods, the total exposure was 3.7 patient‐years for icodec and 0.9 patient‐years for glargine U100. The mean ± SD weekly dose for icodec was 1.69 ± 0.44 U/kg/week (range 1.19–3.08 U/kg/week), corresponding to 0.24 U/kg per day, while the mean ± SD daily dose for glargine U100 was 0.24 ± 0.07 U/kg/day (range 0.15–0.45 U/kg/day). The mean ± SD daily bolus insulin aspart dose over 1 week at steady state was 0.39 ± 0.16 U/kg/day for icodec (following the 7th weekly administration) and 0.43 ± 0.15 U/kg/day for glargine U100 (between the 6th and 13th daily administration). Participants were given additional basal insulin (glargine U100) in the initial 2 weeks of icodec treatment because of the time required to achieve icodec steady state. In Weeks 1 and 2, respectively, participants received 47% and 23% of their individual daily basal dose of glargine U100 determined during the run‐in period (Table S3).

Eleven participants needed very little or no glucose infusion during at least one of the glucose clamps. This was likely due to the clamp target glucose level (6.7 mmol/L) being less than the upper limit of the titration target interval for fasting glucose (4.4–7.2 mmol/L) and due to the operational challenge of achieving appropriate insulin dosing in a clamp experiment with a once‐weekly insulin. Including these participants in the summary of pharmacodynamics was not considered meaningful as they did not meet the pre‐defined necessary conditions for evaluating the glucose clamp results (i.e., blood glucose close to the target level of 6.7 mmol/L and glucose infusion rate above zero for the vast majority of the glucose clamp until potential consistent blood glucose escape and consequently clamp termination). Therefore, based on visual inspection of the individual glucose infusion rate and blood glucose profiles prior to database lock and statistical analysis, these 11 participants were excluded from the pharmacodynamic analysis set (Figure S2). Demographics for the pharmacodynamic analysis set are provided in Table S4 and are similar to the demographics seen in the full analysis set. For participants in the pharmacodynamic analysis set, the mean ± SD weekly dose for icodec was 1.62 ± 0.28 U/kg/week (range 1.21–2.30 U/kg/week), corresponding to 0.23 U/kg per day, while the mean ± SD daily dose for glargine U100 was 0.23 ± 0.05 U/kg/day (range 0.18–0.34 U/kg/day).

Pharmacokinetics

Measured trough concentrations after the start of once‐weekly icodec treatment are shown in Figure 1a, indicating that the average time to reach icodec clinical steady state (as defined by Rowland & Tozer ³⁰ ) was approximately 2–3 weeks. After steady state was achieved, the mean pharmacokinetic profile of icodec following the last dose suggested full coverage of icodec exposure throughout the dosing interval of 1 week (Figure 1b). Further details on the steady‐state pharmacokinetic characteristics of icodec including dose‐proportionality in Japanese participants in this study are presented in Figure S3 and Table S5. In Figure 3, the dose‐normalized mean pharmacokinetic profile of icodec at steady state in Japanese individuals with T1D from the current study is compared with the profile in White individuals with T1D from a previous study ¹¹ , suggesting comparable pharmacokinetics in Japanese and White individuals with T1D.

Figure 1.

Insulin icodec pharmacokinetics (total albumin‐bound and unbound icodec) in Japanese individuals with type 1 diabetes. (a) Measured insulin icodec trough concentrations at the start of a once‐weekly treatment period. The dashed line represents the clinical steady state threshold ³⁰ . The shaded region indicates the insulin icodec trough concentrations considered to be at clinical steady state. The data are presented as means ± standard error of the mean (N = 24). (b) Mean measured serum insulin icodec concentrations during a one‐week dosing interval at steady state. Insulin icodec was given once weekly at a dose titrated individually (1.19–3.08 U/kg body weight). Error bands represent the standard error of the mean (N = 24). N, number of participants; U, unit.

Figure 3.

Mean measured dose‐normalized serum insulin icodec concentrations during a one‐week dosing interval at steady state in Japanese vs White individuals with type 1 diabetes. Data for White individuals are modified from Hövelmann et al. ¹¹ under the terms of the Creative Commons Attribution‐NonCommercial 4.0 International License (http://creativecommons.org/licenses/by‐nc/4.0/). Insulin icodec was given once weekly at a dose titrated individually (mean ± SD 1.69 ± 0.44 vs 1.91 ± 0.44 U/kg body weight, range 1.19–3.08 vs 1.09–3.33 U/kg body weight for Japanese vs White). Error bands represent the standard error of the mean (N = 24 for Japanese and N = 65 for White). N, number of participants; SD, standard deviation; U, unit.

Pharmacodynamics

The partial glucose infusion rate profiles for icodec measured during the glucose clamps as well as the model‐predicted glucose infusion rate profile over the entire dosing interval of 1 week at steady state are presented in Figure 2a, which also shows the physiological fluctuations in insulin sensitivity across the day ³¹ . The model‐predicted daily percentages of the icodec glucose‐lowering effect over the course of a week are illustrated in Figure 2b. Over the span of 1 week, the daily percentages of the full weekly glucose‐lowering effect ranged between 10.7% (on Day 7) and 18.0% (on Day 2). A relative glucose‐lowering effect of 14.3% per day suggests a uniform distribution over the week as evidenced by the dashed line in Figure 2b. Individual blood glucose profiles are presented in Figure S4, and mean measured glucose infusion rate profiles are presented in Figure S5 for the icodec glucose clamps from 24 to 48 h and from 150 to 168 h after the weekly icodec administration and for the glargine U100 glucose clamp from 0 to 24 h after the daily glargine U100 administration.

Figure 2.

Pharmacodynamic effect of insulin icodec during a dosing interval of one week at steady state in Japanese individuals with type 1 diabetes. (a) Mean measured (continuous lines) and model‐predicted (dashed line) glucose infusion rates. Measured profiles are from glucose clamps during a weekly dosing interval at steady state. Insulin icodec was given once weekly at a dose titrated individually (1.21–2.30 U/kg body weight). Error bands represent the standard error of the mean for clamp data (N = 13). (b) Model‐predicted distribution of pharmacodynamic effect over 7 days. The dotted line represents an even distribution of pharmacodynamic effect over these 7 days. The data are presented as means ± standard error of the mean (N = 13). AUC, area under the curve; GIR, glucose infusion rate; N, number of participants; U, unit.

Safety

Throughout the treatment periods, a total of 16 AEs were reported with icodec (4.3 events per PYE) and two AEs were reported with glargine U100 (2.2 events per PYE). All events were mild (15 events for icodec and two events for glargine U100) or moderate (one event for icodec) in severity. The investigator assessed the majority of events to be unlikely related to the study medication. The only AEs reported more than once were nasopharyngitis (three events for icodec) and vessel puncture site pain (2 events for icodec and 1 event for glargine U100). One injection site reaction was reported during icodec treatment and was assessed by the investigator to be possibly related to icodec. No serious AEs or AEs leading to withdrawal were reported. At the end of the study, all AEs were recovered (15 events for icodec and two events for glargine U100) or recovering (one event for icodec).

Throughout the treatment periods (however with the glucose clamps excluded), rates of overall Level 2 hypoglycemia were 37.3 episodes per PYE for icodec vs 30.6 episodes per PYE for glargine U100. Rates of nocturnal Level 2 hypoglycemia were 7.8 vs 9.4 episodes per PYE for icodec and glargine U100, respectively. There were no severe (Level 3) hypoglycemic episodes reported. Rates of hyperglycemia (plasma glucose >16.7 mmol/L) were 33.6 episodes per PYE for icodec vs 36.5 episodes per PYE for glargine U100.

DISCUSSION

The main results of the present study in Japanese participants with T1D were that (a) the time to reach clinical steady state with icodec was approximately 2–3 weeks, (b) the steady‐state icodec pharmacokinetic profile suggested full coverage of exposure throughout the one‐week dosing interval, (c) the proportion of icodec glucose‐lowering effect per day over the week ranged from 10.7% (on Day 7) to 18.0% (on Day 2), and (d) icodec was well‐tolerated.

The pharmacokinetic properties of icodec were previously investigated in White individuals with T1D ¹¹ in a study with essentially the same design as the current study, thereby allowing for comparison between Japanese and White individuals. Since icodec doses were individualized in both studies, the pharmacokinetic profile and exposure‐related end points were dose‐normalized prior to comparison. Figure 3 shows that the one‐week pharmacokinetic profile at steady state was comparable in Japanese and White individuals with T1D both regarding shape and overall exposure level. Likewise, the half‐life of icodec was approximately one week in both populations (Table S5). Icodec pharmacokinetics in White individuals are essentially similar between T1D and T2D ¹⁰ , ¹¹ . By bridging, it is therefore reasonable to assume that the pharmacokinetic properties of icodec are similar in Japanese vs White individuals for both T1D and T2D.

Icodec has a distinct mechanism of protraction and action, which should be carefully considered when interpreting the pharmacokinetic and pharmacodynamic results for icodec. Upon subcutaneous administration, icodec enters the bloodstream and creates a strong, reversible binding to albumin, thereby establishing a primarily inactive reservoir from which icodec is constantly released ⁹ , ¹² . The current pharmacokinetic results are derived from measurement of total serum icodec concentrations, encompassing both albumin‐bound and unbound icodec. However, icodec action on the insulin receptor is mainly exerted by the small fraction of unbound icodec. Furthermore, the affinity of icodec for the insulin receptor is low, suggesting that protraction of icodec glucose‐lowering effect occurs to a significant extent at the level of the insulin receptor, an aspect not captured by the measured icodec levels in the bloodstream ⁹ . The presently available assay technology does not allow measurement of the concentration of unbound icodec in the circulation. It follows that the pharmacokinetic characteristics of icodec may not always fully predict its pharmacodynamic effect. Accordingly, in the current study, the maximum exposure was observed during Day 1, while the maximum glucose‐lowering effect occurred on Day 2. The delay between drug exposure and glucose‐lowering effect usually seen for insulin may also contribute to this observation ³² .

The glucose‐lowering effect profile in Japanese individuals with T1D in the current study (Figure 2a) was overall comparable with the profile previously observed in White individuals with T1D ¹¹ , including approximately the same average glucose infusion rate over the one‐week dosing interval at steady state in both populations. The daily glucose‐lowering effect as a proportion of the total effect over the week ranged from 8.4% on Day 7 to 19.6% on Day 2 in the previous study in White individuals, while the corresponding range in the current study was 10.7% on Day 7 to 18.0% on Day 2. These findings suggest a comparable pharmacodynamic profile of icodec in White and Japanese individuals with T1D. Rates of Level 2 hypoglycemia for icodec were also close to comparable between the two studies, being 32.8 episodes per PYE in White individuals with T1D and 37.3 episodes per PYE in Japanese individuals with T1D in the current study.

The current pharmacological findings support the potential of icodec as a once‐weekly basal insulin. Accordingly, several global Phase 3a trials that also enrolled Japanese participants have shown that icodec can provide basal insulin coverage with once‐weekly administration ¹³ , ¹⁴ , ¹⁷ , ¹⁸ . In basal‐only regimens in participants with T2D, either insulin‐naïve or during basal insulin‐switch, icodec showed noninferiority and statistical superiority in HbA_1c lowering vs once‐daily basal insulin with low rates of clinically significant or severe hypoglycemia in all treatment groups ¹³ , ¹⁴ . During basal‐bolus treatment in T2D, icodec was noninferior vs once‐daily glargine U100 with respect to change in HbA_1c, and the rate of clinically significant or severe hypoglycemia was similar between treatments ¹⁷ . A subgroup analysis of the Japanese participants showed results overall similar to the findings in the global trial populations, thus supporting the use of icodec in Japanese individuals with varying degrees of T2D progression ³³ . Finally, in individuals with T1D, icodec showed noninferiority vs once‐daily degludec with respect to change in HbA_1c during the 26‐week main part of the trial, while the rate of clinically significant or severe hypoglycemia was statistically significantly higher for icodec vs degludec ¹⁸ .

To mimic the use of basal insulin in clinical practice and to ensure safety during the treatment periods, particularly during the 8‐week icodec treatment, the basal insulin dose was titrated individually prior to the treatment periods to achieve SMPG of 4.4–7.2 mmol/L before breakfast. To facilitate a glucose infusion rate above zero in the glucose clamps, the glucose clamp target level was set at 6.7 mmol/L, that is, near the upper range of the titration interval while still being euglycemic. Nonetheless, a total of 11 individuals required minimal or no infusion of glucose throughout at least one clamp experiment and, therefore, the necessary conditions for correct interpretation of the glucose clamp data were not met in these individuals. Consequently, they were excluded from analysis of glucose‐lowering effect. Since the circumstances were similar with icodec and glargine U100, the exclusion of the 11 individuals from analysis of glucose‐lowering effect is deemed unlikely to significantly alter the conclusions drawn from the present study. Still, the number of participants contributing to the pharmacodynamic results was lower than planned, which is a limitation of the present study. Another limitation of the glucose‐lowering effect data on icodec in this study was that direct measurement of glucose infusion rate in a glucose clamp setting occurred only from 24 to 48 h and from 150 to 168 h of the full one‐week dosing interval. The low number of hours spent in a glucose clamp setting was due to logistical constraints and to ensure a manageable burden on the participants. Notably, there was a considerable similarity between the measured partial glucose‐lowering effect curves and the full weekly model‐predicted glucose‐lowering effect curves. Finally, it should be noted that the present results are only fully applicable to Japanese individuals with T1D. Individuals with T1D were chosen for this study to be able to obtain pharmacodynamic results with minimal interference from endogenous insulin. Given the similarity in pharmacokinetic and pharmacodynamic properties of icodec between White and Japanese individuals with T1D, it is reasonable to assume that the pharmacological properties of icodec in Japanese individuals with T2D are comparable to those shown previously in White individuals with T2D ¹⁰ .

In conclusion, the pharmacological characteristics of icodec in Japanese individuals with T1D shown in the current study are comparable to those previously shown in White individuals with T1D. The present findings support the potential of icodec to provide basal insulin coverage with once‐weekly dosing in Japanese individuals with diabetes, thereby reducing the number of basal insulin injections from at least 365 to 52 per year when compared to daily basal insulin treatment.

DISCLOSURE

Niels R. Kristensen, Andrea Navarria, Tomoyuki Nishida, Rasmus Ribel‐Madsen, Stinne B. Søgaard, and Inge B. Halberg are employees and shareholders of Novo Nordisk. Takashi Eto and Miwa Haranaka declare no conflicts of interest.

Approval of the research protocol: The protocol for this study has been approved by a suitably constituted institutional review board (Hakata Clinic IRB, Fukuoka, Japan) and it conforms to the provisions of the Declaration of Helsinki.

Informed consent: All informed consent was obtained from the participants.

Approval date of registry and the registration no. of the study: Clinicaltrials.gov NCT03766854 (04 December 2018).

Animal studies: N/A.

Supporting information

Figure S1. Overall study design.

Figure S2. Participant flowchart.

Figure S3. Relationship between weekly insulin icodec dose and total exposure at steady state in Japanese individuals with type 1 diabetes.

Figure S4. Individual blood glucose profiles during glucose clamps after once‐weekly insulin icodec or once‐daily insulin glargine U100 administration at steady state in Japanese individuals with type 1 diabetes.

Figure S5. Mean glucose infusion rate profiles during glucose clamps after once‐weekly insulin icodec or once‐daily insulin glargine U100 administration at steady state in Japanese individuals with type 1 diabetes.

Table S1. Blood sampling for pharmacokinetic analysis of insulin icodec.

Table S2. Demographics—All participants.

Table S3. Insulin glargine U100 supplementation during the first 2 weeks of the insulin icodec treatment period.

Table S4. Demographics—Pharmacodynamic analysis set.

Table S5. Pharmacokinetic end points in Japanese vs White individuals with type 1 diabetes.