Fourteen clinical trials evaluated the pharmacokinetic profile, efficacy, as well as safety of insulin icodec. Various results included pharmacokinetics, dosing, and outcomes such as change in A1C, continuous glucose monitoring (CGM) parameters, weight gain, and hypoglycemic episodes. Details and results of the studies are shown in Table 1, Fig. 1, and discussion of the trials are summarized below.

Change in A1C Between Insulin Icodec and Active Comparator [16,17,18,19,20,21]

Nishimura et al. [11] evaluated pharmacological and molecular properties of insulin icodec. Their in vitro cell-based studies revealed the when a C20 fatty diacid-containing side chain is added to insulin icodec, it results in strong, reversible binding to albumin. Additionally, three amino acid substitutions (A14E, B16H and B25H) enhance its molecular stability and reduce insulin receptor binding and clearance, thereby extending the half-life. In a clinical pharmacology trial involving subjects with T2DM, insulin icodec was tolerated well and exhibited pharmacokinetic and pharmacodynamic characteristics suitable for once-weekly dosing. The mean half-life was 196 h, and the glucose-lowering effect was distributed evenly over the entire week-long dosing interval. A limitation was the small number of subjects studied (n = 12), however this study provided much value to help confirm the pharmacology and kinetics on insulin icodec.

Pieber et al. [13] compared the timing of the pharmacokinetic profiles of the administration of insulin in an open label study of 46 patients with T2DM. The study showed, in order to reach steady state, 3–4 doses of insulin icodec must be administered. The glucose lowering effects of insulin icodec follows its pharmacokinetic effects in that the most glucose lowering occurs between days 2 and 3 while it slowly tapers down towards the end of the week. However, the overall difference between days is fairly low. This small, open-label study shows insulin icodec can safely be used once a week without great alterations in blood glucose levels.

Plum-Mörschel et al. [10] conducted a randomized, open-label, crossover trial comparing the sites of insulin icodec administration and their pharmacokinetic and pharmacodynamic response. Administration of insulin into the subcutaneous tissue of the abdomen, upper arms, and thigh had similar effects on glucose lowering despite higher peaks when administered in the upper arm or abdomen in comparison to the thigh. Strengths include randomization and both pharmacokinetic (insulin levels) and pharmacodynamic (glucose lowering effect) endpoints. Limitations of the study are its small (n = 25 and open label design.

Bajaj et al. [14] conducted a trial to investigate two methods of switching to insulin icodec from once-daily insulin glargine U-100 in patients with T2DM who were using daily basal insulin and one or more oral glucose-lowering medications. This phase 2, randomized (1:1:1) trial involved basal insulin–treated subjects with T2DM (A1C 7.0–10.0%) switching to icodec with an initial 100% loading dose (where only the first dose was doubled), switching to icodec without a loading dose, or continuing insulin glargine U-100 for 16 weeks. The results showed that switching from daily basal insulin to once-weekly icodec did not exhibit many adverse effects provided effective glucose control in both insulin icodec groups. The rates and incidences of adverse effects and hypoglycemic episodes were similar across the groups. A loading dose at initiation when switching to once-weekly icodec significantly increased the percentage of TIR during weeks 15 and 16 compared to once-daily IGlar U100, without increasing the risk of hypoglycemia. The trial’s limitations included its modest sample size and relatively short duration. However, it had many strengths such as it multicenter and randomized design, low treatment discontinuation rate, and use of CGM.

Lingvay et al. [15] studied the safety and efficacy of insulin icodec using various once-weekly titration algorithms in a phase 2, open-label, randomized, 16-week, treat-to-target study. The trial included adults that were insulin-naive (n = 205) with T2DM and A1C levels of 7–10% who were on oral glucose-lowering medications. Participants began once-weekly icodec and were titrated weekly using different algorithms which varied in FBG goal ranges and dosing increments. The percentage of TIR during weeks 15 and 16 was monitored using CGM continuous glucose monitoring. The study found that once-weekly icodec was well tolerated and effective across the three titration algorithms. The titration algorithm A (FBG goal: 80–130 mg/dL; ± 21 units/week) achieved the greatest balance between glycemic control and hypoglycemia risk. However, titration B (± 28 units/week with the same FBG goal) improved TIR but resulted in more hypoglycemia. The strengths of this study were the multicenter and randomized design, low treatment discontinuation amount, and assessment of glycemic control using TIR with CGM data. ​​The main limitation of this study was its short length of 16 weeks.

Philis-Tsimikas et al. [12] authored a design paper describing the characteristics and rationales of the ONWARDS series. The primary glycemic outcomes measured were the change in A1C from the beginning of the study to week 26, A1C differences from baseline to week 52, change in FBG, and TIR. Additionally, all studies compared safety outcomes, namely hypoglycemia. Insulin icodec was studied in various populations including T2DM (ONWARDS 1–5) [16,17,18,19,20], insulin-naive patients (ONWARDS 1, 3, 5) [16, 18, 20], previously insulin-treated patients (ONWARDS 2, 4) [17, 19], and T1DM (ONWARD 6) [21]. Finally, the series compared insulin icodec to various other insulins including insulin glargine U-100 (ONWARDS 1,4, 5) [16, 19, 20], concentrated insulin glargine U-300 (ONWARDS 5) [20], and long acting insulin degludec (ONWARDS 2, 3, 5, 6) [17, 18, 20, 21]. Overall, this composition of studies allows for broad generalizability for patients with both T1DM and T2DM who either had been on insulin or were insulin naive compared to multiple different basal insulins.

Rosenstock et al. [16] evaluated once-weekly insulin icodec in comparison to once-daily insulin glargine U100 in patients with T2DM that were not previously taking insulin (ONWARDS 1). A1C was reduced in patients taking insulin icodec compared to patients taking insulin glargine and both noninferiority and superiority of insulin icodec were confirmed. Insulin icodec also showed a higher TIR compared to insulin glargine. Hypoglycemic events were not statistically different at 52 weeks but icodec was inferior to glargine at 83 weeks. This is the longest trial studying insulin icodec so far. Strengths were the extended duration of the study, randomized treatment methods which included safety follow-up. Limitations were that this trial did not include a double-blind, double-dummy design and the authors stated the reason was because of the intent to limit the confounding effects on trial subjects from the number of injections that would be needed over the course of the extended duration of the study.

Another Philis-Tsimikas et al. [17] study evaluated insulin icodec to daily insulin degludec in type 2 patients already taking basal insulin (ONWARDS 2). Overall the effect on A1C from baseline to week 26 was greater with insulin icodec than insulin degludec which showed not only a non-inferiority but also superiority of icodec compared to degludec. Weight gain was significantly higher in subjects taking icodec versus degludec and the combined level 2 and 3 hypoglycemia were similar. Compared with degludec, subjects receiving icodec had higher percentages of A1C less than 7% without clinically significant hypoglycemia, and a greater improvement in Diabetes Treatment Satisfaction Questionnaire (DTSQ). The strengths of the study included its randomized, large, multi-center design with diverse participants, high trial completion percentage, use of double-blinded CGM, and gathering of patient feedback on diabetes treatment satisfaction. However, several limitations should be noted. The open-label design could introduce bias into patient-reported parameters and the report of hypoglycemia. Additionally, the trial was powered to evaluate the primary outcome, so any statistically significant differences between groups for secondary outcomes might not be from true clinical effect.

Another study by Lingvay et al. [18] evaluated once-weekly insulin icodec versus once-daily insulin degludec in subjects with T2DM that were not previously taking insulin (ONWARDS 3). A1C was significantly reduced in patients taking insulin icodec compared to patients taking insulin degludec and both noninferiority and superiority of insulin icodec were confirmed. Episodes of hypoglycemia were found to be greater with subjects taking insulin icodec compared to degludec but these values were not statistically significant. There was no discernable difference in weight between the two study groups. The study’s strengths included its double-masked, double-dummy design, and high trial completion rate, which ensured it was well-powered to assess the primary endpoint. However, the study had several limitations. Its 26 weeks duration meant that sustained effects could not be predicted. Additionally, to allow a bases of comparison between the two groups, both insulins were titrated on a weekly basis, whereas conventional adjustment of once-daily degludec could occur daily. Another limitation of the trial was that it was only powered to evaluate the primary outcome, so any statistical differences between the two groups for secondary outcomes, including hypoglycemia, does not necessarily reflect a clinical effect. Lastly, this trial did not use CGM to study TIR as a parameter.

Mathieu et al. [19] evaluated a group of type 2 subjects already taking basal-bolus insulin to assess any differences in being on basal insulin glargine versus basal insulin icodec (ONWARDS 4). This study evaluated adults from 80 sites across nine countries. Overall, both insulin glargine and insulin icodec reduced A1C from baseline to week 26 but was not statistically significantly different and showed a non-inferiority of icodec compared to glargine. All adverse effects were lumped together so it was difficult to assess differences but numerically they appeared to be similar. Also similar were the reported rates of level 2 and level 3 hypoglycemia. The strengths of this study included blinded CGM, high trial completion rate, and inclusion of large, multinational, diverse subjects. However, the trial had some limitations. It was short in duration and had an open-label design. The authors did note that the open-label design was selected for safety reasons, as it was not possible to blind two treatment groups without risking confusion between the two basal pens and the one bolus pen in a double-blind, double-dummy design.

Bajaj et al. [20] also compared a group of insulin-naive subjects with T2DM to assess differences between taking insulin icodec with a predetermined dosing guide app and using daily basal insulin analogues (degludec, glargine U100, glargine U300, selected by study investigator at initial screening) (ONWARDS 5). This study involved adults from seven countries and 176 sites. The results showed that the effect on A1C from baseline to week 52 was higher with insulin icodec using the app than daily basal insulin analogues. This difference was significant, demonstrating not only non-inferiority but also superiority of insulin insulin icodec with app use. Additionally, the DTSQ total treatment satisfaction score for week 52 was significantly higher with icodec and the app compared to daily basal analogues. The Treatment Related Impact Measure for Diabetes (TRIM-D) domain score that measures compliance was also statistically greater for icodec with app. Strengths of this study were its broad inclusion criteria, investigator discretion on choice and dose of daily basal analogues, and individualized trial site visit schedules which reflect clinical practice in the real-world. Limitations were plenty. It had an open-label trial design which could have brought in bias on how the trial was conducted. The authors also note that although it was a 52 week trial, it was not long enough to evaluate diabetes-related and cardiovascular disease outcomes related to the use of insulin icodec.

Russel-Jones et al. [21] evaluated the use of once-weekly insulin icodec compared to once-daily insulin degludec in T1DM participants as part of a basal-bolus set up in a 52 week randomized, open label phase 3a trial across 12 countries (ONWARD6). Adults with T1DM with A1C < 10.0% were randomized to (1:1) to once-weekly icodec or once-daily degludec, both which were in combination with insulin aspart (two or more daily injections). Primary endpoints were change in A1C from baseline to week 26 and test for non-inferiority. The results of the study showed non-inferiority of icodec to degludec. Strengths of the study were its multicentered design, inclusion of patients with T1DM, use of CGM monitors, and long duration. Limitations of the study included open-label design which could lead to risk of potential bias and the utilization of self-measured blood glucose readings compared to CGM for insulin dose adjustments.

Rosenstock et al. [22] also conducted a 78-week randomized, open-label, treat-to-target phase 3a trial in adults with T2DM (A1C 7- 11%) who had not previously been given insulin. Participants were randomized to a 1:1 ratio to receive once-weekly insulin icodec or once-daily insulin glargine U-100. The primary endpoint were the differences in A1C from baseline to week 52. The total mean reduction in A1C at 52 weeks was higher with icodec compared to with insulin glargine, which confirms the noninferiority and superiority of icodec. Overall the rates of hypoglycemic episodes appeared to be less than one event per person-year of exposure for the entire trial. This trial did have blinding of glucose measurements and a long duration, however it had a low proportion of Black and Latino participants.

Pieber et al. [23] also conducted a randomized, open-label, cross-sectional study comparing hypoglycemia in insulin icodec and insulin glargine treated subjects. The study involved a run in period to calculate a dose for the patients. The patients were then administered both double and triple their normal dose for icodec and for glargine. Next, patients were monitored to compare hypoglycemia trends. Once hypoglycemic, the patients were administered intravenous glucose to achieve euglycemia. The proportion of patients achieving clinically significant hypoglycemia (BG < 54 mg/dL) in the two groups were similar. However, patients receiving triple dosing of glargine achieved higher levels of severe hypoglycemia (BG < 45 mg/dL) compared to those receiving triple doses of icodec. Additionally, those taking triple doses of glargine had lower nadir blood glucose compared to icodec. Alternatively, those taking double doses of icodec took longer and needed more glucose to recover from hypoglycemia compared to insulin glargine. Time to achieve clinically significant hypoglycemia was similar between the two groups. Counter-regulatory hormones (namely adrenaline and cortisol) were greater in the triple dose of icodec vs glargine. Finally, symptoms (mental status and vitals) changed in a similar fashion between the two groups. Although small and open-label, this study concluded that similar hypoglycemic reactions occur between icodec and glargine.