In December 2025, the U.S. Food and Drug Administration (FDA) qualified the first artificial intelligence (AI)–powered drug development tool for use in clinical trials for metabolic dysfunction-associated steatohepatitis (MASH).1 The AI-Based Histologic Measurement of NASH (AIM-NASH) tool was qualified under the FDA’s Drug Development Tool (DDT) program, marking a significant regulatory milestone for the use of AI-assisted pathology in later-stage clinical development.
For sponsors and contract research organizations (CROs), this qualification offers a practical way to reduce inefficiency and risk driven by histologic variability in MASH trials.
Why Histology Remains a Structural Constraint in MASH Trials
Assessing MASH is a challenging task in clinical research. Despite advances in non-invasive biomarkers and imaging modalities, regulatory decision-making relies on histologic endpoints from liver biopsies. These endpoints are central to eligibility determination, interim analyses, and efficacy assessment in registrational programs.
However, histologic assessments are a structural bottleneck. Conventional workflows require multiple independent expert pathologists to score biopsy slides using the NASH Clinical Research Network framework. This process is labor-intensive, time-consuming, and subject to inter- and intra-reader variability.2
The research implications include extended cycle times, higher operational costs, and increased uncertainty in endpoint interpretation. Variability at the histology level can obscure treatment effects, increase sample size requirements, and introduce risk to statistical confidence and regulatory interactions.2
Against this backdrop, the FDA’s qualification of AIM-NASH is an opportunity to re-examine how histologic endpoints are generated, standardized, and governed without compromising regulatory rigor or human oversight.
NOTE: Nonalcoholic steatohepatitis (NASH) and metabolic dysfunction–associated steatohepatitis (MASH) refer to the same progressive inflammatory liver disease, with MASH reflecting updated nomenclature that emphasizes the underlying metabolic dysfunction rather than exclusion of alcohol use.
What AIM-NASH Is and Is Not
AIM-NASH is a cloud-based, AI-assisted histology measurement tool that supports pathologists in the evaluation of digitized liver biopsy slides.3 The system applies machine-learning algorithms to quantify four histologic features central to MASH assessment:
- steatosis
- hepatocellular ballooning
- lobular inflammation
- fibrosis
These measurements align with the NASH Clinical Research Network scoring framework that underpins regulatory endpoints in MASH trials.3
AIM-NASH functions as decision support rather than as an autonomous diagnostic. Pathologists retain responsibility for final interpretation and scoring. The tool is qualified for a defined context of use in MASH drug development. It does not replace sponsor obligations related to data integrity, risk-based monitoring, or compliance with Good Clinical Practice.
Why FDA Qualification Matters and What It Signals for Future MASH Programs
The FDA’s qualification of AIM-NASH under the DDT program allows the tool to be used across MASH development programs without requiring sponsors to re-submit the same validation data for each trial. The initial qualification (i.e., supported by validation studies demonstrating high concordance between AI-assisted measurements and expert consensus readings, as well as reproducibility across sites and readers) provides FDA confidence that, when used as intended, the tool produces reliable and reproducible measurements suitable for regulatory decision-making.1,3
What This Means in Practice for Sponsors and CROs
1. Efficiency Gains
AI-assisted scoring reduces the manual burden of histologic review by enabling a streamlined, single-reviewer model supported by algorithmic quantification. Even modest reductions in review time can accelerate eligibility decisions, interim analyses, database lock, and overall trial velocity.
2. Standardization Across Sites
AI-assisted scoring reduces histologic review burden by enabling a streamlined, algorithm-supported single-reviewer model, accelerating eligibility decisions, interim analyses, and database lock.
3. Optimized Resource Allocation
With expert pathology resources scarce, AI-assisted workflows allow sponsors and CROs to redeploy pathologists to higher-value activities such as complex adjudication, biomarker integration, and quality oversight, while preserving human accountability.
A Framework for Integrating AIM-NASH into MASH Trial Operations
The FDA’s qualification of AIM-NASH reflects growing regulatory acceptance of AI tools with defined contexts of use. For sponsors and CROs, the value is realized when AI-assisted histology is embedded in existing pathology, data, and governance workflows rather than deployed as a standalone technology.
The checklist below is designed to help sponsors and CROs operationalize AIM-NASH to deliver measurable benefits.
Workflow and Review Model
- Redesign histology review workflows to incorporate AIM-NASH outputs in a streamlined, single-reviewer model.
- Define clear escalation and adjudication pathways for complex or borderline cases requiring expert judgment.
Data Integration and Infrastructure
- Integrate AIM-NASH outputs with downstream data flows, enabling interpretation alongside biomarker, imaging, and clinical data.
- Establish interoperability with electronic data capture, clinical trial management systems, and analytics platforms.
- Assess digital pathology infrastructure and cloud security controls to support compliant data exchange and storage.
Quality Oversight and Monitoring
- Update quality oversight processes to monitor consistency across sites, readers, and timepoints.
- Leverage AI-assisted trend analysis to identify variability and potential quality risks early.
- Define quantitative metrics to track adoption, review cycle time, and data consistency.
Governance and Accountability
- Update standard operating procedures to reflect AI-assisted histology workflows.
- Implement audit trails and version control for AIM-NASH outputs and human review decisions.
- Document human accountability, with pathologists retaining final responsibility.
Regulatory Readiness and Compliance
- Confirm alignment with AIM-NASH’s FDA-qualified context of use.
- Incorporate AI deployment into trial-specific, risk-based data quality plans.
- Ensure regulatory documentation is audit-ready across studies and sites.
Training and Change Management
- Establish training programs for pathologists, study teams, and monitors on AIM-NASH use and limitations.
- Define change-management plans to support consistent adoption across sites.
AIM-NASH should be evaluated based on program-level impact, not technical performance alone; failure to achieve efficiency and consistency gains indicates an integration issue, not a technology issue.
Key Insight
AI-assisted histology adds value when embedded within site networks, data workflows, and governance models designed to reduce variability, accelerate decisions, and preserve data integrity.
At Alliance Clinical Network, our site strategies support consistent, high-quality data in clinical trials through:
- Prevalence-aligned site planning to reduce endpoint variability.
- Academic and community networks to improve biopsy and data consistency.
- Patient-centered processes to reduce biopsy burden and support retention.
- Early performance tracking to identify variability and operational risk.
1 U.S. Food and Drug Administration. (2025, December 8). FDA qualifies first AI drug development tool, will be used in MASH clinical trials.
2 Ratziu V, Hompesch M, Petitjean M, et al. Artificial intelligence-assisted digital pathology for non-alcoholic steatohepatitis: current status and future directions. J Hepatol. 2024 Feb;80(2):335-351.
3 Pulaski, H., Harrison, S.A., Mehta, S.S. et al. Clinical validation of an AI-based pathology tool for scoring of metabolic dysfunction-associated steatohepatitis. Nat Med 31, 315–322 (2025).
