Skip to content
Buy This GLP
← Research
Evidence

IcoSema: two trials, 98% heterogeneity, and every benefit a surrogate

A weekly injection combining insulin icodec with semaglutide. The two pooled trials disagree almost completely on glucose, and no cardiovascular event was counted.

Neil Sanders7 min read
Two trials, 1,970 people, pooledHbA1c−0.37% (−0.95 to 0.21)body weight−6.10 kg (−7.21 to −5.00)systolic pressure−2.45 mmHg (−3.52 to −1.38)LDL cholesterolratio 0.94 (0.90 to 0.98)triglyceridesratio 0.94 (0.91 to 0.97)cardiovascular eventsnot measuredHeterogeneity on the HbA1c pool: I² = 98%.The two trials agree almost not at all on the outcome they were built for.

Two studies went into this pool. They disagree with each other by 98%.

What I² of 98 means

Heterogeneity measures how much the variation between studies exceeds what chance would produce. At 98%, almost none of the difference between these two trials is random. [1]

Averaging them anyway gives −0.37% on HbA1c with an interval from −0.95 to 0.21 — a wide null. The honest reading is not that the combination fails to control glucose. It is that these two trials measured something different enough that one number cannot stand for both.

One compared against weekly insulin icodec alone and the other against full basal-bolus therapy. Those are different comparators, which is reason enough for the answers to diverge, and it is the same difficulty as comparing by inference.

What did move

Weight fell 6.10 kg against insulin-based intensification, interval 7.21 to 5.00. That is a large and consistent difference, and it is what you would expect from adding semaglutide to a regimen whose alternative is more insulin.

Systolic pressure fell 2.45 mmHg. Total cholesterol, LDL, triglycerides and VLDL all fell, with ratios between 0.94 and 0.97 — improvements of a size that the widest observational view also reports across cardiometabolic measures.

The appendix nobody should quote

The paper includes an exploratory model estimating what these marker changes might do to cardiovascular risk. Its authors describe it as hypothetical and intended as an illustration only, and they put it in a supplementary appendix.

We mention it so that anyone who meets a number from it elsewhere knows what they are looking at. A risk figure derived from surrogate movements is arithmetic about assumptions, not evidence about people, the way a number needed to treat built from counted events is.

What it would mean if it worked

One injection a week instead of several a day, with less weight and a lower chance of hypoglycemia than insulin intensification. For somebody on basal-bolus therapy that is a genuine improvement in how a life is lived.

It is also the same trade as swapping mealtime insulin for a GLP-1 — fewer injections and less weight, bought at a price nobody in either analysis calculated.

Frequently asked

What is IcoSema?
A fixed-ratio combination of weekly basal insulin icodec and semaglutide in a single weekly injection, studied in type 2 diabetes inadequately controlled on basal insulin.
Does it control blood sugar better than insulin?
The pooled difference was −0.37% with an interval crossing zero and heterogeneity of 98%, meaning the two trials disagree too completely for one number to represent both.
Does it reduce cardiovascular risk?
Unknown. Weight, blood pressure and lipids all improved, but those are risk markers and no cardiovascular event was counted. The authors call dedicated outcome trials essential.
What about the risk model in the paper?
Its authors describe it as hypothetical and an illustration only, and placed it in a supplementary appendix. It is arithmetic about assumptions rather than evidence about people.

Sources

  1. [1] Alper A, et al. (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. PMID 42681675

Where to get it

Price the injectable sellers

The desk lists every seller that publishes an injectable figure, with the advertised price struck against the one a buyer is billed.

Open the price desk

More in Evidence