Hepta, a liquid biopsy company developing blood tests for chronic disease, today announced clinical data showing its blood test more accurately identifies metabolic dysfunction-associated steatohepatitis (MASH) patients eligible for treatment than the three non-invasive tests recommended by current clinical guidelines: FIB-4, ELF, and FibroScan. The 528-patient, multi-cohort study will be presented at the European Association for the Study of the Liver (EASL) 2026 Congress in Barcelona on May 30.
The data will be presented by Jörn Schattenberg, M.D., Director of the Department of Gastroenterology, Hepatology, Endocrinology and Nutritional Medicine at Saarland University Medical Center in Germany, at EASL 2026 as Abstract OS-099. The work was selected for oral presentation and chosen for Best of EASL 2026, which highlights the most clinically and scientifically impactful data, including late-breaking results, presented at the annual Congress.
Rezdiffra and Wegovy are now approved for MASH with moderate-to-advanced fibrosis, stages F2 and F3, but identifying these patients at scale has remained challenging. MASH, a progressive form of fatty liver disease characterized by liver inflammation and damage linked to obesity and metabolic dysfunction, typically produces no symptoms until liver damage is advanced. Liver biopsy, the gold standard, is invasive, resource-intensive, and impractical for routine clinical use. The non-invasive blood tests commonly used today were built to identify F3 and F4 fibrosis, the more advanced stages, and miss nearly one-third of the patients now eligible for therapy. Imaging-based testing further depends on specialized equipment and trained operators, putting it out of reach for the population scale this disease requires. These gaps point to the need for an accurate, scalable, blood-based test that can identify treatment-eligible patients in routine clinical practice.
The study included 528 biopsy-confirmed MASLD patients across three independent cohorts: Duke University, the University Medical Center Mainz, and the screening population of Akero Therapeutics’ SYNCHRONY, a contemporary Phase 3 program in MASH. The dataset is one of the largest multi-cohort cell-free DNA (cfDNA) methylation studies in MASH to date. Hepta’s test analyzes methylation patterns, epigenetic marks on DNA fragments shed from the liver and other organs into the bloodstream, giving clinicians a molecular readout of disease biology from a standard blood draw.
Hepta’s LiquidTransformer platform, a large AI model with more than 100 million parameters designed for cfDNA methylation data, was trained on samples from Duke and Mainz cohorts (n = 349) and validated on the SYNCHRONY cohort (n = 179). In head-to-head comparisons on the hold-out validation set, the platform outperformed FIB-4, ELF, and FibroScan in detecting significant fibrosis (F2+) with an AUC of 0.83. It also outperformed all three tests in separating treatment-eligible F2/F3 patients from earlier-stage and cirrhotic disease, using a Pareto frontiers comparison method. A separate meta-analysis showed the Hepta score outperformed published performance benchmarks for a wide range of non-invasive tests, including all three in clinical guidelines.
The hold-out validation cohort was collected across more than 50 sites and is completely independent of the training cohort. By design, this rules out overfitting to the training data, a central concern for clinical machine-learning diagnostics, and supports confident generalization of the results to new patients and populations.
“Existing non-invasive tests for MASH predate effective therapies and were never optimized for treatment selection,” said Hamed Amini, Ph.D., CEO and co-founder of Hepta. “We built a classifier for the clinically actionable question of identifying treatment-eligible patients. On hold-out validation, head-to-head on the same samples, it outperforms FIB-4, ELF, and FibroScan. Our test identifies adults with noncirrhotic MASH with moderate to advanced liver fibrosis, that is F2-F3, the patient population for whom Rezdiffra and Wegovy are now approved.”
The findings extend a growing body of evidence for Hepta’s LiquidTransformer, its purpose-built AI platform that captures tissue biology from cfDNA to enable application of liquid biopsy in chronic disease. At Digestive Disease Week 2026, the company presented data showing the same cfDNA methylation approach can identify semaglutide weight-loss responders before treatment begins and track drug-induced biological changes over time, extending its capabilities from disease characterization to treatment response. Data presented at AASLD 2025 demonstrated that the platform reads disease biology at the pathway level, establishing the foundational science for both diagnostic and monitoring applications.
“The ultimate blood test for MASH needs to do three things: detect treatment-eligible disease, guide therapy selection, and monitor patient response over time to signal whether a therapy is working,” said Soheil Damangir, Ph.D., CTO and co-founder of Hepta. “This data establishes that we can identify the patients who should be on therapy more accurately than the tests in current guidelines. Our recent DDW data demonstrated the second and third: predicting treatment response before therapy begins and tracking MOA-specific drug effects from blood. Together, this is a platform built to serve the continuum of biomarker needs in MASH.”
About Hepta
Hepta is building a blood-based epigenetic measurement platform for chronic disease, starting with MASH. Founded by former Illumina, GRAIL, and Google scientists, the company is backed by Felicis Ventures, Illumina Ventures, SeaX Ventures, Alumni Ventures, and AME Cloud Ventures. Hepta’s liquid biopsy-native AI platform, LiquidTransformer, decodes cfDNA methylation data from a standard blood draw, enabling tissue-level biological insight without the need for a biopsy. For more information, visit hepta.bio.
View source version on businesswire.com: https://www.businesswire.com/news/home/20260527522804/en/
Media gallery
