Metabolic liver disease and biopsy assessment
Metabolic dysfunction-associated steatohepatitis (MASH) is a fatty liver disease linked to obesity and type 2 diabetes, and trials of new treatments for it measure it from liver biopsies whose readings specialists may not always agree on. In March 2025 the European Medicines Agency's human medicines committee qualified an artificial-intelligence method of reading those biopsies, which means the committee can accept evidence the method generates as scientifically valid in future applications.
What MASH is
MASH, formerly known as non-alcoholic steatohepatitis (NASH), is a build-up of fat in the liver that causes inflammation, irritation and scarring over time, without significant alcohol use or another reason for liver injury. It is linked to obesity, type 2 diabetes, high blood pressure, abnormal cholesterol and abdominal fat, and if untreated it can lead to advanced liver disease.
Biopsy assessment in MASH trials
Testing new MASH treatments often relies on liver biopsies, in which small pieces of liver tissue are taken to confirm inflammation and scarring. These biopsies are the gold standard for demonstrating the efficacy of a new, investigational medicine. The disease activity these trials measure is inflammation and fibrosis. But high variability is a challenge in MASH trials, because the specialists who review the samples may not always agree on how severe the inflammation or scarring is.
AIM-NASH and its qualification
On 20 March 2025, following a public consultation, the EMA's human medicines committee (CHMP) issued its first qualification opinion on an AI-based development methodology: AIM-NASH, a tool that helps pathologists analyse liver biopsy scans to identify the severity of MASH in clinical trials. The tool's machine-learning model was trained on more than 100,000 annotations from 59 pathologists who assessed over 5000 liver biopsies across nine large clinical trials. The qualified tool is locked, which means the model cannot be modified or replaced. CHMP encourages optimisation of the model, while acknowledging that major changes may require the tool to be qualified again.
The evidence submitted to CHMP showed that AIM-NASH readings verified by one expert pathologist can reliably determine disease activity with less variability than the current trial standard, a consensus of three independent pathologists. CHMP agreed that the tool can increase the reproducibility and repeatability of assessments for new MASH treatments. Qualification means that CHMP can accept evidence generated by the tool as scientifically valid in future applications. EMA presents the qualification as a significant step towards integrating AI in medicine development.
The tool is expected to enhance the reliability and efficiency of trials of new MASH treatments by reducing variability in measuring disease activity. It can help researchers obtain clearer evidence of new treatments' benefits from trials that include fewer patients, and so can ultimately bring effective treatments to patients faster.