REIMAGINE 4 misses its dual primary endpoint against tirzepatide

Novo Nordisk reported 68-week results from its REIMAGINE 4 trial on 4 August 2026, tucked into its second-quarter financial report rather than released as standalone trial data ^[1][2]. In the head-to-head study, CagriSema did not beat or match tirzepatide 15 mg on both of the measures the trial was designed around, so the study did not succeed on its own stated terms ^[1].

What the numbers showed

REIMAGINE 4 was an open-label trial that enrolled roughly 1,000 adults with type 2 diabetes whose blood sugar was inadequately controlled on metformin, with or without an SGLT2 inhibitor ^[1]. It compared CagriSema — Novo Nordisk's once-weekly combination of the GLP-1 semaglutide and the amylin analogue cagrilintide — against tirzepatide at the 15 mg dose, the highest approved strength of the molecule sold as Mounjaro and Zepbound ^[1].

At 68 weeks, management reported 15.2% weight loss and a 1.9 percentage point reduction in HbA1c for CagriSema, both on what statisticians call the efficacy estimand ^[1]. The comparator figures for tirzepatide 15 mg were 15.8% weight loss and a 2.2 percentage point HbA1c reduction ^[1].

The trial had a dual primary endpoint, meaning CagriSema had to clear a non-inferiority bar on both weight and blood sugar. It met non-inferiority on weight loss but missed it on HbA1c, so the overall trial failed ^[1].

A few technical points matter when reading those numbers. The trial was open-label, meaning participants and investigators knew which drug was being given. The efficacy estimand estimates what happens if people stay on treatment as intended, which typically produces larger numbers than an analysis that counts everyone regardless of whether they stopped the drug. And because the results came in an earnings disclosure, the detailed safety, tolerability and discontinuation data that usually accompany a full presentation are not yet public ^[1].

Why it matters for patients

For people with type 2 diabetes weighing options, this is one of the few direct comparisons between two next-generation injectables rather than the usual indirect cross-trial math. The headline takeaway is narrow: on the two numbers reported, CagriSema and tirzepatide 15 mg landed close together on weight, while tirzepatide came out ahead on HbA1c by 0.3 percentage points ^[1].

It is also worth being precise about who was studied. REIMAGINE 4 enrolled adults with type 2 diabetes, not people with obesity alone ^[1]. Weight-loss percentages in diabetes trials are generally lower than in obesity trials of the same drugs, and results at 68 weeks cannot be compared directly to trials that ran for a different length of time ^[1].

CagriSema is not yet approved in the United States. Novo Nordisk filed with the FDA in December 2025 for weight management and has said it expects a decision by late 2026 ^[1]. This diabetes readout does not by itself change that application, and the sources do not say whether Novo Nordisk will pursue a separate type 2 diabetes submission. Nothing here changes the availability or labeling of currently marketed semaglutide or tirzepatide products.

The broader competitive picture is shifting quickly. Novo Nordisk also reported that Wegovy pill prescriptions in the US exceeded 265,000 in the week ending 17 July, with more than 5 million prescriptions since launch, and that total weekly Wegovy prescriptions were around 575,000 ^[2]. Eli Lilly's oral GLP-1 orforglipron, branded Foundayo, was approved by the FDA for obesity in April 2026 and authorized in the UK on 10 August 2026 ^[1].

What happens next

The FDA decision on CagriSema for weight management is the nearest milestone, guided to late 2026 ^[1]. Novo Nordisk has also said a higher-dose CagriSema trial is expected to begin dosing in the second half of 2026 ^[1], which suggests the company is still testing whether more drug closes the gap.

Further details from REIMAGINE 4 — including side effect rates, how many participants stopped treatment, and results at doses other than those reported — have not been published in the sources available. Whether the full data set will be presented at a medical meeting or in a peer-reviewed journal, and on what timeline, is not yet known.