The One-Shot Gene Therapy That Wants to Replace Weekly Semaglutide: Fractyl's Rejuva, Explained

We've been tracking the frontier where metabolic medicine meets genetic medicine — the one-time base-editing shot that permanently lowers cholesterol, and the way GLP-1 drugs keep spilling into new indications. Fractyl Health's Rejuva sits at the most ambitious corner of that map. The pitch is simple and audacious: instead of injecting semaglutide or tirzepatide every week for the rest of your life, what if a single procedure reprogrammed your own pancreas to make GLP-1 on demand — permanently?

That's the idea behind RJVA-001, Fractyl's lead gene-therapy candidate. In May 2026 it cleared a real regulatory bar and became, in the company's words, the first AAV gene therapy authorized to enter clinical testing for type 2 diabetes. It's a genuinely novel answer to the biggest unsolved problem of the GLP-1 era. It is also — and we'll be blunt about this throughout — almost entirely unproven in humans. Here's the honest breakdown.

The disclaimer that matters most: Rejuva is an experimental biotech program, not a product. You cannot buy it, and nothing here is medical or investment advice. Every efficacy number below comes from Fractyl's own press releases and conference posters, in mice and large animals — there is no peer-reviewed human data on the gene therapy, and as of this writing no patient has been dosed. We flag the evidence tier on everything.

The problem Rejuva is trying to solve

Weekly GLP-1 injections work remarkably well. Semaglutide drives roughly 15% weight loss in trials, tirzepatide around 20–22%, and retatrutide as much as 24–28%. But they share a structural weakness: the moment you stop, the weight comes back. Discontinuation studies show near-universal regain, because the drug only works while it's in your system. That leaves patients facing an indefinite commitment — a weekly shot, forever, at a cost that compounds — plus the adherence drop-off and the lean-mass questions that come with chronic dosing.

Fractyl's bet is that the cleanest answer to "you have to take it forever" is "what if you only took it once." A one-time therapy would, in theory, erase the adherence problem, the rebound problem, and the recurring cost in a single stroke. The catch is that "once" is a very high bar to clear safely — which turns out to be the whole story of this drug.

What RJVA-001 actually is

Fractyl calls the platform Rejuva and the approach "Smart GLP-1." Mechanically, RJVA-001 is an AAV-based gene therapy — it uses an adeno-associated virus, the workhorse delivery vehicle of modern gene therapy, to carry a gene for human GLP-1 into the cells of the pancreas. [1][2]

Here's the plain-language version. Instead of injecting the GLP-1 drug, you deliver the instructions for making GLP-1 into pancreatic beta cells — the same glucose-sensing cells that make insulin. Those cells become a long-term, built-in GLP-1 "factory." And because beta cells naturally sense blood sugar, the pitch is that secretion is nutrient-responsive: it rises with a meal and settles between them, the way the natural hormone behaves — rather than the flat, always-on systemic level you get from a weekly drug. [2][3]

Two design choices are where the engineering bet really lives:

• Local, not systemic, delivery. Most AAV gene therapies are infused into the bloodstream and end up concentrated in the liver — which is also where AAV toxicity tends to show up. RJVA-001 is instead delivered locally to the pancreas through an endoscopic, ultrasound-guided catheter, depositing the vector into the splenic lobe (roughly the body and tail of the pancreas). Fractyl's large-animal data report the vector concentrating in the targeted pancreatic tissue while largely sparing the liver — about 7 vector copies per nucleus in the pancreas versus under 0.2 in the liver — with the procedure running under 20 minutes in pigs. [3][4] (Company-reported, preclinical.)
• Near-physiologic GLP-1 levels. In mouse studies Fractyl reports circulating active GLP-1 of roughly 10–20 pM with RJVA-001 versus 50–150 pM for pharmacologic GLP-1 drugs — more than five times lower, which the company likens to post-gastric-bypass levels and argues should mean less nausea and vomiting than high-dose injectables. [3] (Company-reported, mouse data — a hypothesis, not a clinical finding.)

A second candidate, RJVA-002, applies the same pancreatic-gene-therapy idea to a GIP/GLP-1 dual agonist for obesity. It's far earlier — preclinical — and not the one entering the clinic. [5]

The preclinical data — and exactly how much to trust it

This is where honesty matters most, because Fractyl's headline numbers are striking and easy to over-read.

• The "beats semaglutide" datapoint. In diabetic (db/db) mice, a single dose of the GLP-1 gene therapy produced up to 50% blood-sugar lowering and 11% weight loss over four weeks, versus 32% glucose lowering and 2% weight loss for chronic semaglutide in the same model. [2] That's the figure you'll see quoted everywhere — and it's a four-week mouse experiment. Mouse-to-human translation in metabolism is historically poor, and a model is not a patient.
• Durability — the actual point. At ObesityWeek 2024, Fractyl presented a 13-week diet-induced-obesity mouse study, its longest data, reporting durable weight and glucose effects and, importantly, that the gene therapy prevented weight and blood-sugar rebound after semaglutide was withdrawn. [5][6] That "off-ramp after the injectable" framing is the heart of the whole thesis. It is also, again, mice over weeks — not humans over years.
• The low-side-effect argument. At ASGCT 2025, the company reported the durable potency plus the low systemic GLP-1 exposure described above — the basis for its "efficacy without the high-dose GI burden" claim. [3]

None of this has been published in a peer-reviewed journal, and none of it is independently replicated. Treat every percentage as company-reported preclinical signal, not as a prediction of what happens in people.

Where RJVA-001 actually is right now

Here's the current, verifiable status — the part that separates a press release from a result:

• May 11, 2026: Fractyl announced Clinical Trial Application (CTA) authorization in the Netherlands to begin the first-in-human study of RJVA-001 in type 2 diabetes — making it, the company says, the first AAV-based gene therapy to enter clinical development for type 2 diabetes. [1][8] That's a real, externally-gated milestone: a European regulator reviewed the package and cleared it into the clinic.
• It's a Phase 1/2, open-label, single-ascending-dose study in adults whose type 2 diabetes is poorly controlled despite multiple agents including GLP-1 drugs — a high-unmet-need, GLP-1-failure population, not treatment-naïve patients. The endpoints are safety, tolerability, and preliminary efficacy. [1]
• No patient has been dosed yet. First-in-human dosing — and the first preliminary human data — are guided for the second half of 2026, subject to site activation. Authorization is not the same as dosing, and an early single-ascending-dose study speaks to safety long before it says anything about durable weight or glucose control. [1]
• There is no U.S. IND. Fractyl is running this ex-U.S. first (the Netherlands, with an Australia application also in progress), not through the FDA as lead regulator. [1]

For context, Rejuva is the second durability play at Fractyl. The company's other program, Revita, is a one-time endoscopic device procedure that resurfaces the lining of the duodenum, aimed at weight maintenance after GLP-1 drugs — and unlike the gene therapy, Revita actually has human data and is Fractyl's nearer-term catalyst. [7] The two share a thesis — "make the metabolic benefit durable" — but they're different technologies at very different stages. Don't let Revita's human results rub off on Rejuva's mouse results.

How it compares to a weekly GLP-1 shot

	RJVA-001 (Rejuva gene therapy)	Weekly injectable GLP-1s
Dosing	One-time endoscopic procedure (aspirational)	Weekly self-injection, indefinitely
How it works	Gene delivered to pancreas → cells make native, nutrient-responsive GLP-1	Exogenous receptor agonist, flat systemic level
GLP-1 exposure	Near-physiologic (10–20 pM in mice)	High systemic (~50–150 pM equivalent)
Stop or adjust?	No — you can't titrate down or switch it off	Yes — stop or dose-adjust anytime
Weight regain on stopping	Designed to prevent it (mouse data only)	Near-universal regain
Human efficacy	Unknown — not yet dosed	Established: sema ~15%, tirz ~20–22%, reta ~24–28%
Stage	Preclinical; Phase 1/2 just authorized	Approved, years of real-world use

The single most important row is "stop or adjust." A weekly drug's biggest practical virtue is that it's reversible — if you don't tolerate it, you stop. A gene therapy gives that up by design, which brings us to the part the slides spend less time on.

The honest bear case

A one-time GLP-1 therapy is a genuinely exciting idea. It's also stacked with risks that are specific to being a gene therapy:

1. Irreversibility — the core structural risk. There is no off-switch. If a patient over-responds, develops an adverse effect, becomes pregnant, needs major surgery, or simply wants to stop, you can't un-deliver the gene. The titratability that makes injectable GLP-1s forgiving is exactly what this approach trades away.
2. AAV immunogenicity and the one-shot ceiling. Many people carry pre-existing antibodies to AAV, and the body mounts a response to the vector after dosing — which can blunt the effect and, critically, block re-dosing. If the single dose under-delivers, there may be no second attempt.
3. Durability is unproven in humans. The animal data run weeks to months. Whether transgene expression — and the benefit — lasts years in people is the entire value proposition, and it's completely untested.
4. The procedure isn't a pen. Endoscopic delivery into the pancreas requires a trained specialist and carries procedural risk (the pancreas is an unforgiving organ). That's a meaningfully higher bar than handing someone an auto-injector, and it complicates the "at scale" story.
5. Manufacturing and cost. AAV production is expensive and complex. A one-time obesity/diabetes gene therapy would have to justify a large up-front price against GLP-1 drugs that are effective, improving, and getting cheaper.
6. A high regulatory bar by definition. Regulators have no precedent for a gene therapy in a non-fatal, already-well-served indication. Precisely because excellent reversible drugs exist, the benefit/risk bar for an irreversible one will be steep.

Where this fits in the GLP-1 arms race

Step back and the contrast is the story. The rest of the field is racing toward more convenient chronic dosing — oral GLP-1s are arriving (Lilly's orforglipron at roughly 11% weight loss, oral semaglutide around 13–14% in its latest trials), next-gen injectables like retatrutide push efficacy higher, and amylin combinations are close behind. The whole industry is making the weekly shot easier to take.

Rejuva bets in the opposite direction: not making the dose more convenient, but eliminating the need to dose at all. No other AAV or cell therapy is anywhere near the clinic for type 2 diabetes or obesity, which makes RJVA-001 a genuine outlier — and its novelty is inseparable from its risk. It's the highest-risk, highest-reward seat in the room.

The bottom line

Fractyl's Rejuva is going after the right problem — the rebound-and-adherence trap that defines the GLP-1 era — with a legitimately clever mechanism and a real regulatory milestone now on the board. That deserves attention.

But keep the evidence tiers straight. Everything that makes Rejuva exciting is, today, a mouse result and a hypothesis. The human study has been authorized, not run; the first data are months away and will be about safety, not durable weight loss; and the irreversibility that makes a "one-shot" cure appealing is the same thing that makes it dangerous if it goes wrong. A 50%-in-mice headline is not a 50%-in-humans promise — and anyone who tells you a one-time GLP-1 cure is around the corner is reading a press release, not a clinical result.

We'll update this post when the first-in-human data read out. Until then, the weekly shot — semaglutide, tirzepatide, retatrutide — remains the only GLP-1 with the one thing Rejuva can't yet claim: proof that it works in people.

---

PeptidePrices tracks the research-peptide market and the science around it. We don't sell peptides, and we have no financial relationship with Fractyl Health. This article is educational and is not medical or investment advice.

Sources

1. Fractyl Health — "Fractyl Health Authorized to Initiate First-in-Human Trial of RJVA-001 in the Netherlands" (May 11, 2026) — https://www.globenewswire.com/news-release/2026/05/11/3291742/0/en/Fractyl-Health-Authorized-to-Initiate-First-in-Human-Trial-of-RJVA-001-in-the-Netherlands-First-Gene-Therapy-Candidate-to-Enter-Clinical-Development-for-Type-2-Diabetes.html
2. Fractyl Health — "Fractyl Health Announces New Results from its Rejuva Platform" (db/db vs semaglutide; Mar 12, 2024) — https://ir.fractyl.com/news-releases/news-release-details/fractyl-health-announces-new-results-its-rejuvar-platform
3. Fractyl Health — "New Rejuva Smart GLP-1 Pancreatic Gene Therapy Preclinical Data ... at ASGCT 2025" (low systemic GLP-1, biodistribution; May 19, 2025) — https://www.biospace.com/press-releases/fractyl-health-unveils-new-rejuva-smart-glp-1-pancreatic-gene-therapy-preclinical-data-highlighting-durable-potency-and-safety-with-limited-systemic-glp-1-exposure-at-asgct-2025
4. Fractyl Health — preclinical local-delivery / biodistribution data (WCIRDC; Dec 12, 2024) — https://ir.fractyl.com/news-releases/news-release-details/fractyl-health-unveils-promising-preclinical-data-demonstrating
5. Fractyl Health — ObesityWeek 2024 durability data + RJVA-002 nomination (Nov 4, 2024) — https://www.globenewswire.com/news-release/2024/11/04/2974429/0/en/Fractyl-Health-to-Present-New-Preclinical-Data-on-Sustained-Weight-Maintenance-and-Blood-Sugar-from-its-Rejuva-RJVA-001-Single-Administration-GLP-1-Pancreatic-Gene-Therapy-Candidate-at-ObesityWeek-2024.html
6. Fierce Biotech — "When Wegovy stops and weight gain starts, could a GLP-1 gene therapy change that?" — https://www.fiercebiotech.com/research/when-wegovy-stops-weight-gain-starts-could-glp-1-gene-therapy-change
7. Fractyl Health — First-quarter 2026 results & business update (Revita REMAIN-1, runway; May 12, 2026) — https://ir.fractyl.com/news-releases/news-release-details/fractyl-health-reports-first-quarter-2026-financial-results-and
8. Clinical Trials Arena — "Fractyl Health's GLP-1 gene therapy trial given green light" (May 2026) — https://www.clinicaltrialsarena.com/news/fractyl-healths-glp-1-gene-therapy-trial-given-green-light/