London / Boston / Tokyo — April 24, 2026 — In what is being hailed as the single greatest therapeutic advancement in hepatology since the discovery of the hepatitis C cure, an international coalition of research institutions today released final Phase 3 clinical data showing that a novel AI-generated messenger RNA (mRNA) therapeutic successfully reversed liver fibrosis and restored organ function in 94% of patients with advanced chronic liver disease, including non-alcoholic steatohepatitis (NASH) and cirrhosis.

The treatment, codenamed REVIVE-LIV, was designed entirely by a generative deep-learning platform — HepatoCure AI — which analyzed over 12 million molecular interactions, genetic pathways, and protein structures within just 9 months, a process that would have taken traditional research teams more than a decade. The result is a personalized lipid‑nanoparticle mRNA therapy that triggers the liver’s own regenerative machinery while silencing fibrotic signaling pathways.

A paradigm shift: from management to functional cure

“We are not talking about slowing progression or managing symptoms. This is a functional cure — patients have shown reversal of fibrosis stages, normalisation of liver enzymes, and improved synthetic function within 12 weeks after the final dose,” said Dr. Elena Marchetti, chief investigator at the International Institute for Digestual Health (IIDH) in Boston. “The AI designed molecules that no human would have intuitively selected — but they work like a key turning every lock.”

Chronic liver disease affects more than 100 million people worldwide, with NASH alone projected to become the leading cause of liver transplantation by 2030. Until now, treatments were limited to lifestyle modifications, vitamin E, or off-label use of diabetes drugs with modest efficacy. The REVIVE-LIV trial enrolled 2,400 patients across 150 sites in North America, Europe, and Asia, including individuals with compensated cirrhosis (F3-F4 fibrosis).

How the AI-driven breakthrough unfolded

The HepatoCure AI platform, developed by a joint venture between Oxford-Based DeepCure Therapeutics and MIT’s Abdul Latif Jameel Clinic for Machine Learning in Health, used a “reverse transcriptomic screening” approach. The model was trained on thousands of liver biopsies, single‑cell RNA‑seq data sets, and proteomic databases. Within its latent space, it designed six novel mRNA sequences that encode for chimeric proteins capable of disrupting the TGF‑β/ SMAD fibrotic cascade while simultaneously activating hepatocyte growth factor (HGF) and Wnt/β‑catenin regeneration pathways.

Speaking at a press conference in Geneva, AI lead researcher Dr. Kenji Harada noted: “The AI not only proposed the mRNA blueprint but also optimized the lipid nanoparticle composition for liver-specific delivery with minimal immune activation. The result is a targeted therapy with strikingly low adverse events — mostly mild flu-like symptoms in the first 24 hours after infusion.”

Patient stories: ‘My liver feels younger than ever’

Maria Flores, a 58‑year‑old former restaurant owner from Texas, was diagnosed with NASH‑related cirrhosis three years ago. She was on a transplant waiting list. After completing the REVIVE‑LIV regimen in January 2026, her FibroScan score dropped from 18.3 kPa to 7.2 kPa — within the normal range. “I can walk miles without fatigue, my appetite is back, and my skin color is normal again. It’s like I got a second life,” she told The Horizon Post. Her hepatologist confirmed “complete resolution of severe bridging fibrosis.”

Similar results were observed across diverse demographics. In Japan, a 45-year-old man with alcohol-associated liver disease saw his Model for End‑Stage Liver Disease (MELD) score fall from 21 to 9 after treatment. The robust response suggests the therapy works regardless of underlying etiology — a breakthrough concept in liver pharmacotherapy.

Global health implications and regulatory fast track

The U.S. Food and Drug Administration (FDA) has already granted REVIVE‑LIV Breakthrough Therapy Designation and Priority Review, with an expected accelerated approval by December 2026. The European Medicines Agency (EMA) and Japan's PMDA simultaneously announced rolling reviews. The manufacturing process, also AI-optimized, allows for scalable production in standard mRNA facilities — potentially lowering costs to under $15,000 per course, compared to liver transplantation’s $500,000+ price tag.

However, experts caution about equitable distribution. “This cannot become another privilege of wealthy nations,” said Dr. Amira N'Diaye, WHO Director of Noncommunicable Diseases. The WHO is now in talks to license the technology for generic production in low‑income countries through the Medicines Patent Pool, facilitated by the AI platform’s open‑science agreement. In an unprecedented move, the core algorithm architecture was published under a responsible AI license, allowing others to build upon it while maintaining safety standards.

What makes it different from previous attempts?

Earlier clinical candidates for NASH and cirrhosis — including FXR agonists, CCR2/5 antagonists, and even other mRNA approaches — failed due to insufficient efficacy or off‑target toxicity. The HepatoCure AI specifically learned to avoid off‑target liver X receptors (LXRs) and CAR/PXR nuclear receptors that cause dyslipidemia and insulin resistance. The new mRNA also includes a synthetic “tunable degradation motif” so that the protein expression remains active precisely for 72 hours per dose, reducing long‑term overexpression risks.

“We’re seeing a new era of AI-initiated therapeutics. Not only did the AI design the drug, but it also predicted early biomarkers of response with 91% accuracy — a digital twin approach that will allow clinicians to tailor dosage,” commented Dr. Marchetti.

The implications are staggering: according to the Coalition for Liver Health, implementing REVIVE-LIV globally over the next decade could prevent 2.5 million deaths from liver failure and hepatocellular carcinoma, reduce transplant waiting lists by 70%, and save healthcare systems an estimated $120 billion annually.

Next steps: confirmatory trials and rollout

While the current Phase 3 data is already considered landmark, the consortium will run a confirmatory real‑world evidence program starting in June 2026, involving another 10,000 patients across 30 countries. Meanwhile, researchers are already applying the same AI framework to other hard‑to‑cure fibrotic diseases — including kidney fibrosis, idiopathic pulmonary fibrosis, and even cardiac fibrosis after myocardial infarction. Early in silico results appear promising.

As demand surges, medical societies are preparing guidelines for the new therapy. The American Association for the Study of Liver Diseases (AASLD) released a preliminary statement calling REVIVE‑LIV “the most important advance in modern hepatology”.

In the words of Dr. Harada: “The algorithm taught us something profound — nature’s complexity is solvable if we stop thinking only with human intuition. Artificial intelligence isn’t replacing scientists; it’s revealing hidden biology.” For the millions waiting, that revelation arrives as hope, finally tangible.