A liver biopsy is an 18-gauge needle passed through the abdominal wall into an organ that sits an inch from the diaphragm. Sponsors have been using it as a primary endpoint in MASLD trials for decades, not because it is ideal, but because no alternative had cleared the regulatory bar. That calculus is shifting, and the signal that moved it just appeared in Nature Medicine.
The study draws on the LITMUS (Liver Investigation: Testing Marker Utility in Steatohepatitis) registry, a prospectively recruited European cohort that now exceeds 10,000 people with MASLD. At that scale, non-invasive tests for liver fibrosis staging are not being validated against a convenience sample. They are being stress-tested against a population-level signal. The findings matter for every sponsor running a Phase 2 or Phase 3 MASH trial with a histological primary endpoint.
The Biopsy Assumption Is Breaking Down
The old way: sponsors designed MASLD trials around liver biopsy as the gold standard for fibrosis staging, accepted sampling variability as a known cost, and built sample sizes large enough to absorb it. The FDA required histological improvement as the primary endpoint for full approval. The burden was treated as fixed.
The new way, or at least the emerging evidentiary case for it: non-invasive tests, specifically vibration-controlled transient elastography (VCTE) and validated serum biomarker panels, now carry enough diagnostic precision to function as surrogate endpoints in well-defined patient subgroups. Published meta-analytic data shows VCTE achieving 84.8% pooled sensitivity and 87.5% pooled specificity for cirrhosis detection (F4), with a pooled diagnostic odds ratio of 41.8. For F3 (severe fibrosis), sensitivity runs at 81.9% and specificity at 84.7%. Those numbers are not perfect, but they are no longer embarrassing next to a biopsy procedure that samples 1/50,000th of the liver’s volume and carries inter-reader variability that can shift a patient’s fibrosis stage by a full grade.
What the LITMUS dataset adds is prospective, real-world validation at a scale that no single-sponsor study could generate.
The regulatory infrastructure is catching up to the science. In a significant move, the FDA’s Center for Drug Evaluation and Research accepted a Letter of Intent for the qualification of liver stiffness measurement by VCTE as a reasonably likely surrogate endpoint in adults with non-cirrhotic MASH with moderate-to-advanced fibrosis. That acceptance does not mean automatic approval for every trial using VCTE. It means the FDA has signaled that the evidentiary pathway exists. The LITMUS paper now supplies a significant portion of the population-level data that pathway requires. On the European side, the EMA’s CHMP issued a Qualification Opinion on an AI-based tool for diagnosing inflammatory liver disease from biopsy images — the agency’s first qualification opinion grounded in machine learning methodology. Both regulators are moving toward the same destination from different directions.
Who Gets Disrupted First
Consider the sponsor running a proof-of-concept Phase 2 in MASLD with fibrosis stage F2-F3 as the target population. Under the biopsy-dependent protocol, enrollment requires histological confirmation at screening, which means a biopsy before the patient receives a single dose. Sites in community gastroenterology practices frequently lack the procedural volume to run that at scale. Patient refusal rates at the screening biopsy step can reach 20-30% in some trial networks, which inflates both screen failure costs and recruitment timelines.
If validated non-invasive tests can replace that screening biopsy, the protocol changes materially. Site eligibility expands. Enrollment accelerates. The patient population becomes less self-selected toward individuals willing to undergo an invasive procedure, which improves the generalizability of the data — exactly the real-world evidence argument the FDA’s 2016 framework, mandated by the 21st Century Cures Act, was designed to support.
The MASLD patient population is enormous. More than 38% of the world’s adult population meets diagnostic criteria, and prevalence among people with type 2 diabetes exceeds 70% in the U.S. The commercial opportunity is driving trial volume fast. Resmetirom (Rezdiffra), a thyroid hormone receptor beta agonist, received FDA accelerated approval in March 2024 as the first approved pharmacotherapy for MASH with liver fibrosis. That approval opened the competitive field. The sponsors now chasing confirmatory data and label expansions face a choice: replicate resmetirom’s biopsy-heavy trial architecture, or design protocols around the non-invasive endpoint framework the LITMUS data is building.
Sponsors in oncology and rare disease have spent the last decade learning that endpoint selection is strategy, not just science. MASLD sponsors are about to learn the same lesson.
The Counterintuitive Risk in This Signal
Here is the assumption most protocol teams will carry into their next design meeting: validated non-invasive endpoints mean simpler trials. Fewer biopsies, lower site burden, faster enrollment, cleaner data. That assumption deserves pressure.
The LITMUS registry was built on a prospectively recruited, carefully characterized cohort with rigorous imaging and laboratory protocols standardized across sites. Reproducing that standardization in a multicenter, multinational Phase 3 trial is operationally demanding. VCTE performance degrades in patients with obesity, which is precisely the population enriched in MASLD. Inter-device variability between FibroScan platforms at different sites introduces measurement noise that the LITMUS dataset controlled for in ways most sponsors will not replicate by default. If a sponsor drops VCTE into a Phase 3 protocol without the same calibration infrastructure LITMUS used, the resulting data may not satisfy the FDA’s qualification criteria even though the sponsor technically used the “right” test.
The FDA’s accepted Letter of Intent for VCTE qualification sets a context: non-cirrhotic MASH, moderate-to-advanced fibrosis. Sponsors running trials in early fibrosis (F1-F2) or in populations with cirrhosis are not operating inside that qualified context. Using VCTE as a primary endpoint outside those boundaries is a different regulatory conversation, and the LITMUS data, however robust, does not automatically extend the evidentiary base to those subgroups.
What to Do Before Your Next Protocol Lock
If you are designing or amending a MASLD trial protocol in the next six months, do three things before you lock the endpoint section. First, map your target fibrosis stage against the FDA’s accepted VCTE qualification context: F2-F3, non-cirrhotic MASH, adult population. If your trial sits within that boundary, you now have population-level evidentiary support from LITMUS to anchor a Type B meeting request for non-invasive primary endpoints. Second, audit your planned site network for VCTE device standardization. The LITMUS registry’s signal strength came from prospective, protocol-driven imaging — not clinic-standard practice. Your site initiation visits need to treat VCTE calibration with the same rigor as ECG equipment qualification. Third, if you are retaining a screening biopsy for any subgroup, document in the protocol why VCTE is insufficient for that specific subgroup and cite the Nature Medicine data explicitly. That documentation will matter during FDA review.
The Forward Look
The FDA’s VCTE qualification process is still open. The Letter of Intent acceptance is not a qualification opinion — it is a declaration that the agency will evaluate the full submission. Watch for the FDA’s formal qualification decision, which would convert VCTE from a “reasonably likely” surrogate candidate into a qualified surrogate endpoint with defined context of use. That decision will land like resmetirom’s approval did in March 2024: quietly for a week, then loudly for the next eighteen months as every competing sponsor recalculates their protocol assumptions simultaneously. The LITMUS paper is the most important piece of supporting evidence for that qualification decision published this year. Start reading it as a regulatory document, not a scientific one.
References
- Nature Medicine — “Non-invasive tests for the evaluation of liver fibrosis in MASLD”
- Innovative Health Initiative — LITMUS Project Factsheet: 10,000+ patient MASLD registry
- FDA — “FDA Accepts Proposal for Reasonably Likely Surrogate Endpoint in MASH”: VCTE Qualification Letter of Intent Acceptance
- Gastroenterology Research — Transient Elastography Pooled Sensitivity and Specificity Data (F3/F4)
- EMA CHMP — Qualification Opinion: AI-Based Measurement of NASH/MASH Histology from Liver Biopsies
- Diabetes Care — MASLD Prevalence: 38% of World’s Adult Population; >70% in U.S. Type 2 Diabetes Patients
- University of Illinois Chicago Drug Information Group — Resmetirom (Rezdiffra) FDA Accelerated Approval, March 2024
- FDA — Framework for FDA’s Real-World Evidence Program (21st Century Cures Act Mandate)
Moe Alsumidaie is Chief Editor of The Clinical Trial Vanguard. Moe holds decades of experience in the clinical trials industry. Moe also serves as Head of Research at CliniBiz and Chief Data Scientist at Annex Clinical Corporation.
