The European Association for the Study of Obesity updated its living obesity-pharmacotherapy framework ↗ on Tuesday, presented at the European Congress on Obesity in Istanbul. Two recommendations changed. When weight loss is the goal and the patient has no other complications, tirzepatide ↗ is now the first-line choice over semaglutide ↗. And for the first time, only one of the two drugs gets the nod for improvement in liver fibrosis stage: semaglutide.
For seven months, the framework had been the obesity field's first major society guideline to organize itself by complication rather than by weight target. The October 2025 original version ↗, also in Nature Medicine, said either drug for uncomplicated obesity. The May 2026 update integrated six newly eligible randomized trials and shifted the evidence enough that tirzepatide alone now carries the recommendation when nothing else is on the chart.
The MASH split
The other move is more substantive. The October version was built on evidence current to January 31, 2025. That was before the ESSENCE trial ↗ of semaglutide 2.4 mg in patients with biopsy-confirmed metabolic dysfunction-associated steatohepatitis at fibrosis stage F2 or F3, which read out in NEJM on April 30, 2025 and hit both primary endpoints: 62.9 percent of the semaglutide arm had steatohepatitis resolution without fibrosis worsening (vs 34.3 percent on placebo), and 36.8 percent achieved fibrosis-stage improvement without steatohepatitis worsening (vs 22.4 percent on placebo). Tirzepatide already had MASH-resolution data from SYNERGY-NASH. So the new framework allows either drug for MASH remission, but specifies semaglutide alone when the question is fibrosis-stage improvement. Tirzepatide does not have a clean fibrosis-staging readout yet.
That distinction matters because most clinical decisions in MASH are not about whether the steatohepatitis is gone. They are about whether the fibrosis is moving in the right direction. Fibrosis stage is the strongest predictor of liver-related mortality and progression to cirrhosis, and fibrosis improvement was one of the two surrogate endpoints the FDA accepted in 2024 for the first accelerated approval ↗ in MASH, granted to resmetirom. A guideline that says "either for resolution, only semaglutide for fibrosis" is in effect telling clinicians that semaglutide is the more durable answer for the patients whose livers are scarring fastest.
What stayed
The rest of the algorithm is unchanged from October. Tirzepatide remains the first pick for obesity with obstructive sleep apnea, where the SURMOUNT-OSA Phase 3 readouts gave it data semaglutide does not have. Semaglutide remains first for obesity with knee osteoarthritis, on the basis of the STEP 9 pain-reduction signal. Either drug for type 2 diabetes or prediabetes. Semaglutide alone for established cardiovascular disease, on the strength of SELECT for major adverse cardiovascular events. Either for heart failure, where tirzepatide has SUMMIT and semaglutide has STEP-HFpEF.
When modest weight loss is the target, the framework still lists liraglutide ↗, naltrexone-bupropion, and phentermine-topiramate. None of those moved.
The evidence base behind this update is sixty-two randomized trials, which the companion systematic review and meta-analysis ↗ lists in full. The cutoff date moved from January 31, 2025 to November 21, 2025. The framework is now formally a living document, with the EASO leads committing to revisit at six-month intervals.
The commercial split
The commercial consequence is asymmetric. Lilly's tirzepatide gains the headline indication. Most prescribers who reach for an incretin without a specific comorbidity in mind will now have a society document pointing them at Mounjaro or Zepbound. Novo's semaglutide keeps the higher-value comorbidities. Cardiovascular disease and MASH-fibrosis are the two indications where prescribers feel like they are picking the drug for a specific reason, and both are semaglutide's. Knee osteoarthritis is a smaller pool but real. The mix of "default for the average patient" against "specialist pick for the harder patients" is a defensible split for both companies, and it is unusual to see a society guideline make the split this cleanly.
The methodological story is worth its own paragraph. Most society guidelines update on a multi-year cycle and lag the evidence by definition. The American Diabetes Association still publishes its Standards of Care annually. The European Society of Cardiology revises its heart-failure guidelines on a roughly four-year cycle. Seven months between framework drops is a different operating tempo. Andreea Ciudin at Vall d'Hebron and Barbara McGowan at Guy's and St Thomas's set up the structure so that any major new RCT can trigger a rewrite. Whether other obesity societies copy the cadence is the next question.
For peptidemodel, the cleanest readout is that the case for picking incretins by complication, not by potency, is now the official European position. The platform's GLP-1R ↗ and GIPR ↗ target pages already separate the candidate set that way; the framework is the first time a society guideline does. The next major readouts are mazdutide GLORY-2 and the head-to-head DREAMS-3, both expected at ADA 2026 in New Orleans on June 5 to 8. Either could trigger another EASO revision under the six-month rule.