Novel siRNA Therapy Significantly Reduces Liver Fat in MASLD Phase 1 Trial

A single injection of an experimental genetic therapy has normalized liver fat in a majority of patients with metabolic liver disease, sustaining the effect for more than eight months. The first-in-human trial results offer a glimpse at a twice-yearly intervention for a condition that currently requires daily or weekly medication.[1][2]

Published in The New England Journal of Medicine and presented at the American Association for the Study of Liver Diseases 2026 meeting, the data demonstrate that the drug CIDEBsiRNA cut liver fat by up to 80 percent. The phase 1 trial tracked 40 adults over 36 weeks.[1][6]

The findings represent a potential paradigm shift for treating metabolic dysfunction-associated steatotic liver disease. While recently approved therapies demand continuous administration to maintain their benefits, this small interfering RNA approach silences the underlying genetic machinery for months at a time.[1][2]

"What is particularly striking in these results is that we actually normalized liver fat in four out of five participants by week 24," said Jörg Täubel, the trial's lead investigator. Täubel serves as an executive chairman at Richmond Pharmacology in London.[1]

Täubel noted that because the profound fat reductions extend well beyond a six-month period, the therapy lends itself to an infrequent dosing schedule. A twice-yearly injection could significantly improve patient adherence compared to lifelong daily pills.[1]

Mimicking a rare genetic shield

The drug's mechanism traces back to a massive genetic analysis conducted by the Regeneron Genetics Center. Researchers screened sequencing data from over a million individuals to identify natural, inherited protections against chronic liver disease.[2][5]

They discovered that people carrying rare loss-of-function mutations in the CIDEB gene naturally exhibited 22 percent lower liver fat. These individuals also faced significantly lower odds of developing the metabolic disease or progressing to severe cirrhosis compared to the general population.[2][5]

Building on that human genetic validation, Regeneron partnered with Alnylam Pharmaceuticals in 2023 to engineer a targeted therapeutic. The resulting drug, CIDEBsiRNA, is designed to temporarily silence the CIDEB gene in the liver, artificially mimicking the protective mutation in patients who lack it.[4][5]

The RNA interference approach works by intercepting the genetic instructions before they can be translated into disease-causing proteins. By delivering a synthetic RNA strand directly to the liver, the drug degrades the messenger RNA responsible for producing the CIDEB protein.[5]

"The magnitude and durability of the reductions in liver fat observed with CIDEBsiRNA following a single administration are particularly encouraging," said Pushkal Garg, chief medical officer at Alnylam. He noted the data provides an important early signal for RNA interference in metabolic disease.[5]

Garg added that the phase 1 performance represents a critical milestone for the underlying technology. It proves the RNA platform can address broad, population-level metabolic conditions, expanding far beyond the rare genetic disorders it was originally developed to treat.[5]

Unprecedented hepatic fat clearance

The double-blind, placebo-controlled trial enrolled 40 adults diagnosed with the metabolic liver condition. Participants were required to have a body mass index between 30 and 40, alongside a baseline liver fat concentration of at least 8.5 percent to qualify for the study.[2][4]

Thirty participants received a single subcutaneous injection of the therapy at one of four escalating dose levels—15, 60, 200, or 600 milligrams. The remaining ten patients received a placebo, and researchers tracked their hepatic fat using magnetic resonance imaging-proton density fat fraction.[2][4]

By week 36, patients across the active treatment groups saw dose-dependent liver fat reductions ranging from 58 to 80 percent. In stark contrast, the placebo cohort recorded a mere 2.9 percent mean decrease over the identical nine-month observation window.[1][4]

The depth of the clearance exceeded standard clinical benchmarks for the disease. Sixty-two percent of the treated patients saw their liver fat normalize completely, dropping below the healthy 5 percent threshold, compared to zero patients in the placebo group.[1][4]

Furthermore, 93 percent of the CIDEBsiRNA recipients achieved at least a 30 percent relative reduction in liver fat. Clinical meta-analyses have consistently linked this specific reduction threshold to the eventual histological resolution of liver inflammation and fibrosis.[1][4]

Biomarkers and safety profile

Beyond advanced imaging, the targeted therapy rapidly improved circulating blood markers of active liver injury. Mean serum alanine aminotransferase levels among all treated patients fell 32 percent from baseline to week 12, signaling a sharp drop in hepatic stress.[1][2]

By the 36-week mark, participants receiving the highest 600-milligram dose maintained a 36.5 percent drop in the enzyme levels. The placebo group, meanwhile, experienced a 6.6 percent increase in the same liver marker over the course of the trial.[1]

"Prior advances in steatotic liver disease have begun to make a difference, but there is so much more progress to be made to help patients normalize liver fat and reverse liver damage," Täubel said in a statement. He emphasized the need for deep, sustained reductions.[5]

The safety profile appeared highly favorable for an early-stage biologic intervention. Treatment-emergent adverse events occurred in 87 percent of the siRNA group and 90 percent of the placebo group, with the vast majority classified by investigators as mild or moderate.[1][2]

Upper respiratory tract infections were the most common issue, affecting exactly 20 percent of both cohorts. One serious event of severe hyperglycemia occurred in the 200-milligram group, but investigators determined it was unrelated to the drug after the patient was diagnosed with adult-onset type 1 diabetes.[1][2]

Crucially, no patients experienced severe liver enzyme elevations, and there were no reports of nausea or vomiting. Those specific gastrointestinal side effects frequently complicate patient adherence to GLP-1 receptor agonists, which are increasingly prescribed for similar metabolic conditions.[2]

The path to phase 3 trials

The disease, formerly known as non-alcoholic fatty liver disease, affects an estimated 13 million adults in the United States alone. Left unchecked, the progressive accumulation of fat triggers chronic inflammation, eventually leading to irreversible cirrhosis and liver failure.[5]

The commercial and clinical stakes for a successful intervention are massive. Regeneron estimates the United States market for advanced metabolic liver disease treatments could expand from roughly $1 billion today to $15 billion by 2035 as diagnosis rates climb.[5]

Current pharmaceutical options are limited but growing rapidly. Madrigal Pharmaceuticals secured regulatory approval for the daily pill resmetirom in March 2024, and Novo Nordisk's blockbuster semaglutide gained a specific liver indication in August 2025.[1][2]

However, both of those approved therapies require continuous, lifelong administration to keep liver fat at bay. Gary Herman, senior vice president for Regeneron Genetic Medicines, told investors that the siRNA's durable effect could fundamentally alter how hepatologists manage the chronic disease.[2]

Regeneron is now advancing CIDEBsiRNA into a phase 2b trial for fibrotic steatohepatitis and a larger phase 3 safety study for broader metabolic liver disease. The phase 2b study will utilize invasive liver biopsies to confirm whether clearing fat translates to reversed tissue scarring.[1][4]

Final data from the biopsy-driven trial is expected in 2029. Until those results confirm definitive histological benefits, the genetic therapy remains strictly investigational, though its phase 1 performance has established a formidable new benchmark for pharmacological fat reduction.[4]