A once-weekly injection that combines a long-acting basal insulin with semaglutide produced a substantial reduction in body weight, lowered systolic blood pressure and improved atherogenic blood lipids in adults with type 2 diabetes, according to a systematic review and meta-analysis published in Cardiovascular Diabetology – Endocrinology Reports. The analysis pooled data from two large phase 3 trials in the COMBINE program, enrolling 1,970 adults who were inadequately controlled on basal insulin. The treatment, known as IcoSema, is the first fixed-ratio combination of basal insulin icodec and a GLP-1 receptor agonist that can be given just once a week, cutting the number of annual injections from at least 365 to 52.
The meta-analysis, conducted by a team led by Adir Alper and Robert Faillace, followed PRISMA reporting guidelines and was prospectively registered in PROSPERO. The researchers searched MEDLINE/PubMed, Embase, the Cochrane Central Register of Controlled Trials and Web of Science, with a final search date of March 2026, and screened records independently with two blinded reviewers. Of the three trials in the COMBINE program, two met the inclusion criteria: COMBINE 1, led by Mathieu and colleagues, which compared IcoSema with once-weekly insulin icodec, and COMBINE 3, led by Billings and colleagues, which compared IcoSema with basal-bolus therapy consisting of once-daily insulin glargine U100 plus two to four daily injections of insulin aspart. COMBINE 2, which compared IcoSema with semaglutide monotherapy in patients escalating from GLP-1 receptor agonist treatment, was excluded because its non-insulin comparator addresses a fundamentally different clinical question and would have introduced substantial clinical heterogeneity into the pooled estimates.
The headline finding was weight. IcoSema reduced body weight by a pooled mean difference of 6.10 kilograms compared with control regimens, with a 95 percent confidence interval of 7.21 to 5.00 kilograms and a P value below 0.00001. Heterogeneity was moderate to high, with an I-squared of 79 percent, because the weight benefit was larger against basal-bolus therapy in COMBINE 3, where the difference was 6.72 kilograms, than against icodec monotherapy in COMBINE 1, where it was 5.59 kilograms. The authors attribute this pattern partly to the weight-gaining tendency of conventional basal-bolus insulin regimens, which makes the contrast against them more pronounced.
Blood pressure also improved. Systolic blood pressure fell by a pooled mean of 2.45 millimeters of mercury with IcoSema, with a confidence interval of 3.52 to 1.38 and consistent effects across both trials and no detectable heterogeneity. Diastolic blood pressure showed a smaller, non-significant trend in the same direction, at 0.51 millimeters of mercury. On the lipid panel, reported as exponentiated treatment ratios equivalent to ratios of geometric means, total cholesterol fell by roughly 3 percent, while LDL cholesterol, VLDL cholesterol and triglycerides each fell by approximately 6 percent, all statistically significant. HDL cholesterol showed a small, borderline increase of about 1 percent that did not reach conventional significance.
Glycemic control told a more complicated story. The pooled HbA1c difference of 0.37 percentage points in favor of IcoSema was not statistically significant, and the heterogeneity was extreme, with an I-squared of 98 percent. The authors are explicit that this pooled estimate should not be read as a single summary effect. The overall figure conceals two very different comparisons: IcoSema was markedly superior to once-weekly icodec monotherapy, reducing HbA1c by an additional 0.66 percentage points, but was essentially equivalent to intensive basal-bolus therapy, where the difference was only 0.07 percentage points. The authors therefore interpret glycemic efficacy through the trial-specific results rather than the pooled value, noting that IcoSema is superior to icodec alone but comparable to aggressive daily insulin regimens.
The safety data are central to the clinical appeal of the regimen. In COMBINE 1, clinically significant or severe hypoglycemia occurred at 0.21 episodes per person-year with IcoSema versus 0.39 with icodec alone. In COMBINE 3, the contrast was far starker: 0.21 versus 2.23 episodes per person-year against basal-bolus therapy, an estimated rate ratio of 0.09. These low rates were achieved alongside superior or non-inferior glycemic control. As expected for a GLP-1 receptor agonist-containing therapy, gastrointestinal events such as nausea, diarrhea, vomiting and decreased appetite were the most frequent treatment-emergent adverse events, predominantly mild to moderate, occurring mainly during dose escalation and generally transient. The authors did not pool hypoglycemia or gastrointestinal events quantitatively because only two trials with differing comparators and event-reporting conventions were available.
The mechanistic explanation for the cardiometabolic signal rests largely on the semaglutide component. GLP-1 receptor agonists lower systolic blood pressure through natriuresis, improved endothelial function and arterial compliance, reduced vascular inflammation and the hemodynamic consequences of weight loss. Their lipid effects are attributed to reduced hepatic VLDL production, delayed gastric emptying that blunts postprandial lipemia, and weight-loss-mediated improvements in insulin sensitivity. The basal insulin component contributes little independent benefit to blood pressure or lipids, so the authors attribute the observed signal mainly to semaglutide and the weight loss it drives. The class itself has established cardiovascular credentials: meta-analyses of cardiovascular outcome trials show GLP-1 receptor agonists reduce major adverse cardiovascular events by 13 to 14 percent, cardiovascular mortality by 12 to 13 percent and heart failure hospitalization by 11 to 14 percent, with injectable semaglutide showing a 26 percent reduction in SUSTAIN-6 and oral semaglutide a 14 percent reduction in the SOUL trial.
The authors are careful to frame these results as hypothesis-generating rather than definitive. With only two trials contributing to each outcome, the pooled estimates are inherently fragile, between-study variance is poorly estimated, subgroup and meta-regression analyses are not feasible, and funnel plots and asymmetry tests lack the statistical power to be meaningful below ten studies. Sensitivity analyses prespecified in the protocol proved uninformative with so few trials. Both source trials were open-label, which raises the possibility of performance and detection bias, particularly for subjective outcomes, although the authors judge that centrally measured endpoints such as weight, blood pressure and laboratory lipids are less vulnerable. Both trials were also sponsored by Novo Nordisk, which was involved in design, conduct and analysis, a recognized potential source of sponsorship bias, though the meta-analysis itself was investigator-initiated and unfunded. The 52-week follow-up may also be too short to capture long-term cardiometabolic trends or late safety signals.
The findings do not extend to all patient groups. Because COMBINE 2 was excluded, the results do not apply to patients escalating from GLP-1 receptor agonist monotherapy or to those who are insulin-naive and GLP-1-naive. In that excluded trial, IcoSema produced a greater HbA1c reduction than semaglutide monotherapy, 1.35 versus 0.90 percentage points, but was accompanied by modest weight gain of 0.84 kilograms rather than the weight loss seen with semaglutide alone, a pattern that differs from the insulin-comparator setting analyzed here. The authors suggest that a future three-way meta-analysis using comparator-adjusted methods may help bridge this gap. GRADE certainty ranged from low for HbA1c, downgraded for very serious inconsistency, to moderate for all other outcomes.
No dedicated cardiovascular outcome trial for IcoSema has been completed, and to the authors’ knowledge none is yet registered or expected to report in the near term. An exploratory, hypothetical modeling exercise applying the pooled surrogate-marker changes to the 2013 ACC/AHA Pooled Cohort Equations is presented in the supplementary appendix purely as an illustration, not as a clinical prediction, and the authors warn it should not inform clinical decision-making. For now, any expectation of cardiovascular benefit rests on indirect evidence from the semaglutide component rather than on the combination product itself. The authors conclude that once-weekly IcoSema delivers meaningful improvements in weight, blood pressure and atherogenic lipids, the key modifiable drivers of atherosclerotic disease, with reduced injection burden and low hypoglycemia risk, but that dedicated cardiovascular outcome trials are essential to confirm whether these surrogate benefits translate into fewer cardiovascular events.
Subject of Research: Cardiometabolic effects of once-weekly IcoSema in adults with type 2 diabetes
Article Title: Cardiometabolic outcomes of once-weekly IcoSema in adults with type 2 diabetes: systematic review and meta-analysis of the COMBINE trials
Article References: Alper, A., Fagin, A., Karthikeyan, A., Lerner, M., Gershon, A., Morgan, A. A., Rubinstein, G., Agarwal, R., & Faillace, R. (2026). Cardiometabolic outcomes of once-weekly IcoSema in adults with type 2 diabetes: systematic review and meta-analysis of the COMBINE trials. Cardiovascular Diabetology – Endocrinology Reports, 12(1), Article 59. https://doi.org/10.1186/s40842-026-00333-0
Image Credits: AI Generated
DOI: 10.1186/s40842-026-00333-0
Keywords: IcoSema, type 2 diabetes, insulin icodec, semaglutide, meta-analysis, COMBINE trials, body weight, blood pressure, lipids, hypoglycemia, cardiovascular risk, GLP-1 receptor agonist
News Source: Ophelia Keating. (October 28, 2026). Weekly IcoSema injection cuts weight and blood pressure in type 2 diabetes. Scienmag.
