GLP-1 drugs like Eli Lilly's Zepbound and Novo Nordisk's Wegovy work to address obesity—as long as a patient keeps taking them.
Harith Rajagopalan, M.D., Ph.D. thinks his company, Fractyl Health, may have the answer: a gene therapy that integrates a GLP-1-secreting transgene directly into the beta islet cells of the pancreas, locking in the mechanism that gives weight loss drugs their benefits.
“The dialogue has shifted from what amazing things these drugs can do to how we keep those effects when people stop the drug because the rebound is so violent and significant that it has become the unmet need in obesity,” Fractyl CEO Rajagopalan told Fierce Biotech Research in an interview. “We’ve got great drugs that can lower your weight and cardiovascular risk, but if the majority of people stop taking them within months and everything gets worse right away, how are we going to actually solve the healthcare problem of obesity?”
The Fractyl team presented new data June 23 at the American Diabetes Association Scientific Sessions showing that the AAV9-based drug Rejuva is an effective “off-ramp” for weight loss drugs in mice, allowing them to maintain weight loss of 22% post-GLP-1 agonist treatment at eight weeks, the equivalent of about five years in humans.
“Lilly and Novo have fantastic drugs with high potency, and they really prove the GLP-1 mechanism and its benefit, but the product form is limited,” Rajagopalan explained. While other players in the field are trying to develop drugs with longer half-lives, these still don’t dodge the problems of discontinuation and loss of effect, he said.
“I think that we stand alone in trying to develop therapies that address those fundamental limitations in the class,” he said.
Cellular-level differentiation
Fractyl’s ADA conference data comes from a study on 30 mice with obesity. At the start of the experiment, ten were given daily injections of semaglutide, the active ingredient in Wegovy and Ozempic. Another ten received a single administration of Rejuva directly into their pancreas while another ten served as untreated controls.
By four weeks, the cohort that initially received Rejuva had lost 21% of their body fat, while the ones that were given semaglutide lost 16%—results consistent with findings Fractyl shared in October 2023. The weight loss corresponded to improvements in glucose and insulin levels, indicating that the treatments were effective against metabolic disease. The mice given Rejuva also preserved more lean muscle than the ones on semaglutide alone.
At that point, the researchers stopped semaglutide injections, randomized the mice in that cohort and gave five of them Rejuva. Four weeks later, they compared those that were treated with the gene therapy to those that weren’t.
Rejuva effectively kept the grams off in the absence of semaglutide. Mice in the crossover arm maintained a 17% fat loss and a 22% total weight loss, while the ones in the discontinuation group regained all but 1% of their baseline weight. The difference extended to their biomarkers, too: Mice that didn’t cross over to Rejuva lost all improvements in their insulin and glucose levels, while they continued to get better in the mice that did.
But the real proof was in the beta islet cells of the Rejuva-treated mice. When the researchers isolated the cells and treated them with various amounts of glucose, they found that they secreted GLP-1 according to how much glucose was present.
“The reason that matters is because it's the first demonstration that our construct, which uses the insulin promoter, is truly nutrient-responsive in how it secretes GLP-1,” Rajagopalan said. “We think that that is an incredible differentiator of this approach.”
If Rejuva reaches the clinic, it will be administered directly into the pancreas using a special device. The approach may solve a couple of key problems that have come up with gene therapies in the past, the first of which is permanence. Given that beta islet cells don’t divide, one treatment may be enough for long-term efficacy. Second, because the treatment isn’t administered systemically and will likely be effective at relatively low doses, it’s less prone to toxicity than other gene therapies delivered via AAV9, which is already widely considered to be safe, Rajagopalan explained.
He compared Fractyl’s gene therapy to Zolgensma, Novartis’ gene therapy for spinal muscular atrophy. Zolgensma also works on non-dividing cells—neurons—and, like Rejuva, uses an AAV9 vector.
“If you're using a capsid where there's a ton of experience with it already, and you have two to three orders of magnitude less of a dose than what is being used in an approved product, we believe that reduces the risk of AAV-associated toxicity,” he said. That said, whether Fractyl’s therapy will be less effective in people who have antibodies to the adenovirus—an old challenge for gene therapies—is an open question that will require clinical studies to answer, he said.
For the company’s diabetes program, such studies are anticipated for the first half of 2025. Fractyl has already nominated a candidate, Rejuva 001, and is in the process of completing investigational new drug-enabling studies. It will nominate an obesity candidate by the end of 2024, Rajagopalan said.
While it’s too early to talk about price, Rajagopalan acknowledged that gene therapies are expensive. But because Rejuva is delivered locally, it might be less costly than systemic treatments because many fewer particles are needed for it to be effective, he said.