Duodenal resurfacing limited regain after stopping GLP‑1s

What happened

On July 15, 2026, Fractyl Health reported one-year results from the REMAIN-1 Midpoint Cohort, a randomized, double-blind, sham-controlled study of Revita, an investigational endoscopic procedure that resurfaces the duodenal lining (the first stretch of small intestine) in adults with obesity who had lost at least 15% of their body weight on tirzepatide and then stopped the drug. The question the study asks matters to a huge share of GLP-1 patients: when the medicine stops, weight regain almost always follows. Can a single procedure hold the loss without restarting the drug?

The one-year answer in this 45-person pilot: people who got Revita regained an average of 7.8% of their body weight over the year, versus 13.0% in people who got a sham procedure, roughly a 40% relative reduction in regain. People are not regaining nothing; the procedure blunted the rebound, it did not abolish it.

The Midpoint Cohort enrolled 45 adults with obesity who had achieved at least 15% total body weight loss on tirzepatide. After discontinuing the drug, participants were randomized 2:1 to a single Revita procedure (29 people) or a sham endoscopy (16 people), and followed for a year. Neither patients nor assessors knew who got which. Revita is done through the mouth with an endoscope: a balloon catheter ablates (resurfaces) the duodenal lining, with the idea of restoring nutrient-sensing signals disrupted by chronic metabolic disease.

An important caveat built into the design: this cohort was a pilot meant to de-risk the larger Pivotal Cohort, and it was not powered for formal statistical significance. The numbers are descriptive. The company analyzed results with GLP-1 weight loss fitted as a covariate, reporting least-squares means at a 20% run-in-loss reference.

• Full modified intention-to-treat population (N=45): 7.8% regain with Revita versus 13.0% with sham at one year, about 40% less regain.
• People who received a complete duodenal ablation longer than 14 cm (n=17 Revita, 16 sham): kept about 81% of their GLP-1-induced weight loss at one year, versus 48% in sham. Least-squares regain was 4.8% versus 13.0%, more than a 60% relative reduction. This dose-response pattern, longer ablation, more effect, supports the mechanism.
• An "optimized" subgroup with complete ablation plus higher run-in weight loss of at least 17.5% (n=10 versus 8) kept about 84% of the loss versus 46% with sham. Note the shrinking sample sizes: these subgroup numbers rest on 10 people.
• Responder rate: 73% of the mITT group maintained at least 5% total body weight loss relative to their pre-tirzepatide weight at one year, rising to 91% among those with complete ablation. That responder definition is one of two co-primary endpoints in the ongoing Pivotal Cohort, where the FDA set a pre-specified performance goal above 50%.

No device- or procedure-related serious adverse events occurred through one year. No new device-related treatment-emergent adverse events appeared between six and twelve months. Overall treatment-emergent adverse event rates were nearly identical between arms (24% Revita versus 25% sham), a profile the company describes as generally mild and compatible with outpatient use. One new type 2 diabetes diagnosis occurred in the sham arm, none with Revita.

This one-year readout is the third step in a sequence the company has reported. At three months (September 2025), Revita patients had lost an additional 2.5% of body weight after stopping tirzepatide while sham patients had regained 10% (p=0.014). At six months (January 2026), regain was 4.5% versus 7.5% in the pre-specified efficacy population, and exploratory analyses showed improved HDL cholesterol, triglyceride-to-HDL ratio, and reduced sweet-food craving. The one-year data extend the same trajectory: the effect persists and scales with ablation length.

What it means

For anyone who has stopped, or is planning to stop, a GLP-1 because of cost, insurance, side effects, or simple fatigue with injections, these results sketch a possible future option that is not another drug: a one-time endoscopic procedure aimed at keeping the loss. Revita has FDA Breakthrough Device designation specifically for weight maintenance after GLP-1 discontinuation. It is CE marked in the EU and UK.

But the decisive evidence is not in yet. The fully powered REMAIN-1 Pivotal Cohort, which must clear both co-primary endpoints (six-month regain versus sham, and the one-year 5%-loss responder rate above the FDA performance goal), reports topline data in early Q4 2026, with a potential regulatory submission to follow. Those results, in a much larger group, are what would actually support approval.

What it does not mean

It is not an approval, and Revita is investigational only in the United States. No patient can currently buy it here outside a clinical trial.

It is not a reason to stop a working medicine hoping to get Revita. Everyone in this study had already stopped tirzepatide as part of the protocol, and the procedure is not available as standard care.

It is not proof of long-term, multi-year benefit. One year is meaningful, but obesity is chronic; whether a single procedure changes the five-year trajectory is unknown.

And the caveats on the numbers are real: 45 people, a 2:1 randomization, no formal power for significance, subgroup findings resting on as few as 8-10 participants per arm, and company-reported results from a company webcast and press release rather than a peer-reviewed publication. Midpoint data can also shift when the pivotal cohort locks. The 40%-less-regain headline is best read as "promising pilot signal," not "demonstrated treatment effect."