Summary: In 679 adults with type 2 diabetes inadequately controlled on daily basal insulin, once-weekly IcoSema (insulin icodec plus semaglutide) was non-inferior to daily basal-bolus therapy for HbA1c change at week 52 (estimated treatment difference -0.06 percentage points, 95% CI -0.22 to 0.09). On confirmatory secondary endpoints IcoSema was superior, with 6.72 kg greater weight loss, 270 U lower weekly insulin dose, and an 88% lower rate of clinically significant or severe hypoglycaemia. This was an open-label, manufacturer-sponsored phase 3a trial.
PICO Summary
| Element | Detail |
|---|---|
| Population | 679 adults with type 2 diabetes (HbA1c 7.0-10.0%) inadequately controlled on daily basal insulin (20-80 U/day); mean age 59.6 years, 59% male. Open-label, treat-to-target, non-inferiority, randomised, phase 3a trial across 109 sites in 14 countries (NCT05013229). |
| Intervention | Once-weekly IcoSema, a fixed combination of basal insulin icodec and the GLP-1 analogue semaglutide, by pen device (n=340). |
| Comparison | Basal-bolus therapy: once-daily insulin glargine U100 plus two to four daily insulin aspart injections (n=339). |
| Outcome | Primary (HbA1c, week 52): -1.47 pp (IcoSema) vs -1.40 pp (BBT); ETD -0.06 pp (95% CI -0.22 to 0.09), p<0.0001 for non-inferiority (0.3 pp margin) – non-inferior, not superior. Confirmatory secondary (superiority confirmed): bodyweight ETD -6.72 kg (95% CI -7.58 to -5.86, p<0.0001); weekly total insulin dose ETD -270 U (95% CI -303 to -236, p<0.0001); clinically significant or severe hypoglycaemia 0.21 vs 2.23 episodes per person-year (rate ratio 0.12, 95% CI 0.08 to 0.17, p<0.0001). Adverse events: gastrointestinal disorders 148/340 (44%) with IcoSema; infections 116/328 (35%) with BBT. |
COMBINE 3: IcoSema vs basal-bolus
RCT · type 2 diabetes · 52 weeks
Once-weekly IcoSema matched daily basal-bolus on HbA1c reduction and was superior for weight, insulin dose, and hypoglycaemia.
Expert Commentary
COMBINE 3 was designed to test whether a once-weekly combination injection could match the glycaemic ceiling of a daily basal-bolus regimen, and on that primary question the answer is yes: HbA1c reduction was non-inferior, with a treatment difference close to zero. The verdict is that IcoSema is a credible intensification option that simplifies a four-injection-a-day regimen to one weekly injection, and the superiority signals on weight, total insulin requirement, and hypoglycaemia are clinically meaningful rather than marginal. These secondary gains are mechanistically plausible, since substituting bolus insulin with a GLP-1 component is expected to reduce both weight and hypoglycaemia, so the magnitude is not implausibly inflated. The principal limitation is the open-label design, which is difficult to avoid given the divergent injection schedules but leaves treat-to-target titration and patient-reported outcomes vulnerable to expectation bias. The trial was funded by the manufacturer, Novo Nordisk, and three authors are employees, so independent replication of the weight and hypoglycaemia advantages would strengthen confidence. Can I use this with my patients? Not yet, because IcoSema is not a licensed, marketed product in most jurisdictions, but the most likely future candidate is the patient on basal insulin with rising HbA1c, weight gain, or recurrent hypoglycaemia who struggles with daily bolus dosing. Clinicians should track regulatory approval and real-world hypoglycaemia data before adopting this approach.
References
Billings LK, Andreozzi F, Frederiksen M, Gourdy P, Gowda A, Ji L, et al. Once-weekly IcoSema versus multiple daily insulin injections in type 2 diabetes management (COMBINE 3): an open-label, multicentre, treat-to-target, non-inferiority, randomised, phase 3a trial. Lancet Diabetes Endocrinol. 2025;13(7):556-567. doi:10.1016/S2213-8587(25)00052-X
