CagriSema and EloraTZP Both Beat Tirzepatide. Read the Doses

Within nine days in September 2026, both companies that dominate obesity medicine announced a win against the same drug. On September 21, Novo Nordisk said its combination CagriSema beat tirzepatide on weight loss in a phase 3 trial. On September 30, Eli Lilly, which makes tirzepatide, said its own combination, EloraTZP, beat tirzepatide too. Tirzepatide, sold as Mounjaro and Zepbound, has become the yardstick. The two announcements measure very different things.

Why everyone is adding amylin

Both new products pair an incretin drug with a second hormone signal: amylin, which the pancreas releases after meals and which acts on fullness. CagriSema combines Novo's long acting amylin analog cagrilintide with semaglutide, the molecule in Ozempic and Wegovy, in a single weekly injection. EloraTZP pairs Lilly's selective amylin agonist eloralintide with tirzepatide, so it hits three hormone pathways at once: GIP, GLP-1 and amylin. The bet in both cases is that two kinds of satiety signal add up to more weight loss than one.

CagriSema against tirzepatide: a fair fight?

REIMAGINE 5 enrolled 1,023 adults with type 2 diabetes not controlled on metformin, an SGLT2 inhibitor or both, according to its ClinicalTrials.gov record, which lists the trial as open label. Participants got either CagriSema at 1.0 mg of each component or tirzepatide at 5 mg, once a week, for 60 weeks.

At week 60, Novo reports average weight loss of 12.4 percent on CagriSema against 9.1 percent on tirzepatide, a gap of 3.3 percentage points that it calls superior. HbA1c fell 1.71 points on CagriSema and 1.67 on tirzepatide, which met the bar for noninferiority. Both weight figures use the efficacy estimand, which estimates what would happen if everyone had stayed on treatment as assigned.

Now look at the doses. Tirzepatide comes in six strengths, from 2.5 to 15 mg, and 5 mg is the second lowest. CagriSema 1.0/1.0 is also a low strength: in REDEFINE 9, a separate obesity trial announced the same day, Novo tested it alongside a higher 1.7/1.7 version. So REIMAGINE 5 compares a low dose against a low dose. Novo frames that openly, saying not all patients receive or stay on the highest dose. It is a legitimate clinical question. It is not the question most readers assume when they see "beats tirzepatide."

Novo beat tirzepatide at the low end of the dial. Lilly beat its own drug at the top.

What we do not have yet is just as important. The announcement is topline only: no side effect rates, no discontinuation figures, no confidence intervals, no peer reviewed paper. Novo says the drug was well tolerated, which is a description, not a number. In REDEFINE 9, in people with obesity, CagriSema 1.0/1.0 produced 21.0 percent weight loss at week 68 against 2.0 percent on placebo.

EloraTZP against the top dose of Zepbound

Lilly's study is a different animal. It is a phase 2b trial, double blind and placebo controlled, that randomized 367 adults with obesity or overweight and type 2 diabetes in the US and Argentina into ten groups: placebo, eloralintide alone at three doses, tirzepatide alone at its top dose of 15 mg, and several combinations. By my arithmetic, that is fewer than 40 people per group.

From an average starting weight of 105.4 kilograms and an HbA1c of 8.1 percent, the results at 48 weeks, again using the efficacy estimand, looked like this:

• Placebo: 3.0 percent weight loss, HbA1c down 0.3 points.
• Eloralintide alone: 8.2 to 12.3 percent, HbA1c down 1.1 to 1.4.
• Tirzepatide 15 mg alone: 14.8 percent, HbA1c down 2.4.
• Lowest combination (eloralintide 3 mg with tirzepatide 5 mg): 13.2 percent, HbA1c down 2.2.
• Highest combination (eloralintide 9 mg with tirzepatide 15 mg): 23.3 percent, about 24.6 kilograms, HbA1c down 2.9.

At the top, that is 8.5 percentage points beyond tirzepatide at full dose, my subtraction from Lilly's table. Only the comparison of the combination with placebo was the formal primary endpoint. The head to head comparisons with tirzepatide alone were secondary.

The cost: dropouts

This is where the Lilly release deserves a careful read. Stomach side effects were more common in the combination arms, mostly during dose escalation. Discontinuation because of adverse events ranged from 10.8 to 27.0 percent across EloraTZP groups, against 2.9 percent on tirzepatide alone. In the worst arm, roughly one person in four quit because of side effects. Placebo, oddly, came in at 16.7 percent, a figure that in groups this small can move a lot with a handful of people.

Lilly's answer is a new product for phase 3: a single co formulated injection rather than the two separate shots used here, and an optimized escalation schedule. Phase 3 is due to start by the end of 2026. Whether a slower ramp keeps the weight loss while cutting the dropouts is the whole question, and nobody can answer it from this trial.

What to watch next

Three dates matter. Novo filed CagriSema with the FDA for weight management in December 2025 and expects a decision in the fourth quarter of 2026, so an approval could arrive before the full REIMAGINE 5 data are public. Lilly's EloraTZP phase 3 program should begin by year end, and eloralintide alone is already in phase 3 for obesity. And the detailed REIMAGINE 5 results, whenever they come, will reveal what the topline leaves out: the tolerability profile. If reading two trials side by side feels like comparing apples to oranges, that is because it usually is; this guide to reading a health study covers why, and the orforglipron versus oral semaglutide duel is a rare case of a true head to head. As always, none of this is a reason to switch or adjust a treatment without a doctor.

I read these two announcements as positioning more than science. Novo needed a sentence with "superior to tirzepatide" in it, and it designed a trial that could deliver one, at the low end of the dose range where its product is most competitive. Lilly, meanwhile, can afford to compare its new combination against its own blockbuster at full strength, because whichever wins, Lilly sells it. The real race is not on weight loss anymore; past 20 percent, an extra few points matter less than how many people can stay on the drug for years. On that measure, the most telling number of the month may be Lilly's 27 percent dropout rate, and the most telling absence is Novo's missing tolerability table.