Europe’s next evidence gatekeeper: why EMA’s Qualification of Novel Methodologies now deserves board-level attention

Qualification of Novel Methodologies (QoNM) has moved beyond its old role as a specialist regulatory procedure. It is becoming one of the clearest signals of how Europe intends to judge AI tools, imaging methods, mechanistic models and other next-generation evidence instruments in drug development. 

For years, the EMA’s Qualification of Novel Methodologies process was easy to leave to specialists. It sat in the part of the regulatory map reserved for biomarkers, technical working groups and long procedural documents.

That is no longer the right way to read it.

The framework has been in place since 2008, and by the 2023 multi-stakeholder workshop EMA reported 208 completed QoNM procedures, including 27 published Qualification Opinions and 39 Letters of Support. What has changed is not simply the count. It is the kind of methods now pressing into the system, and the strategic weight that comes with them. 

The real shift is this: QoNM is becoming one of the places where Europe decides what counts as decision-grade evidence before a medicine ever reaches formal assessment.

That gives the process a much broader significance than its name suggests. It now touches AI pathology, imaging quantification, PBPK modelling, registries and data sources, and even device-bridging strategies. Read that list carefully and the implication becomes obvious.

This is no longer a side conversation about methodology. It is a live conversation about admissible evidence. 

What QoNM actually is

At its core, QoNM is a voluntary scientific pathway through which developers can ask European medicines regulators to qualify an innovative development method within a predefined Context of Use. That phrase matters more than almost anything else in the process.

EMA is not endorsing a technology in the abstract.

It is assessing whether a specific method is acceptable for a specific regulatory purpose. The route can end in confidential Qualification Advice, a public Qualification Opinion, or a public Letter of Support when a methodology is promising but not yet ready for full qualification. 

The procedure is structured, not informal.

Under the applicant guidance, the process begins with IRIS submission around day -60 and a preparatory meeting around day -30. Qualification Advice can be adopted around day 100. Qualification Opinions run longer because they move through public consultation before final CHMP adoption around day 190, with the draft opinion released for six weeks of consultation.

That architecture is one reason QoNM carries more weight than many teams assume: it is built to shape standards, not merely to exchange views. 

Why EMA is rebuilding the framework

The current structure was not designed for a world of living methodologies. That is the pressure now running through the system.

At the 2023 workshop, stakeholders pushed EMA to clarify scope, sharpen how Context of Use is defined, facilitate earlier interactions, and think seriously about lifecycle management, especially for AI/ML methods and evolving registries.

The workshop report is strikingly direct on this point: for AI-based methods, regulators need a clearer way to distinguish changes that require reassessment from changes that can sit within an ongoing quality system.

The same report argues that qualification cannot be treated as the endpoint for dynamic tools such as registries, where validity and fitness-for-purpose must be revisited over time. 

EMA turned those concerns into a September 2024 action plan.

The plan laid out work on updated procedural guidance, new briefing-document and output templates, methodology-specific Q&As, stronger communication, and enhanced early support. It also explicitly proposed lifecycle-management rules for Qualification Opinions, including criteria for re-validation and possible approaches for periodic interaction around qualified registries.

In the December 2025 update to industry, EMA added more detail: scope definitions, early interaction support, and free virtual scoping meetings were all part of the redesign, while the delivery timeline for the broader package was pushed back because the work proved more substantial than expected. 

That delay is not a sign of drift so much as a sign of where the real difficulty sits.

As of March 31, 2026, the main public QoNM page still points readers to the applicant guidance last updated on October 4, 2023, alongside the September 2024 action plan.

In other words, the public framework is visibly in transition: the strategic redesign is underway, but the core public architecture still reflects an earlier generation of the process. 

The opinions that made this real

The public opinions from 2024 to 2026 make the change impossible to ignore.

The turning point came in March 2025, when EMA announced its first Qualification Opinion for an AI-based methodology: AIM-NASH, used to assess disease activity in MASH/NASH biopsy samples in clinical trials. EMA said the tool could reduce variability, improve reproducibility and repeatability, and support clearer evidence generation in future applications.

Just as important, EMA described the qualified model as “locked,” meaning it cannot simply be modified after qualification without consequences; major changes may require re-qualification.

That is exactly the kind of regulatory language that signals a move from enthusiasm about AI to governance of AI. 

The same pattern appears elsewhere. In August 2025, EMA published a Qualification Opinion for Simcyp Simulator, qualifying the platform for predicting the average magnitude of selected CYP-mediated drug-drug interactions within specific contexts of use.

The opinion is tightly bounded: certain CYP pathways are in scope, the qualified use is limited to defined interaction scenarios, and relevant changes to systems, compound models or the qualification matrix may require updated uncertainty analyses or further justification.

Again, the signal is not broad endorsement. It is controlled acceptability inside a clear evidentiary frame. 

In March 2026, EMA published a Qualification Opinion for the Molecule-independent device bridging approach, or MIDBA. The method was qualified as an alternative approach for bridging from manual subcutaneous injection with a handheld or prefilled syringe to an autoinjector for monoclonal antibodies, but only under specified conditions around dose, formulation, injection volume, needle length, injection site and absorption characteristics.

This is a useful reminder that QoNM is not drifting into abstractions. It is being applied to concrete development bottlenecks. 

The 2024 Centiloid opinion points in the same direction from a different angle.

EMA qualified the Centiloid measure of amyloid PET to quantify brain amyloid deposition, and the opinion states that the metric can be used for enrichment in clinical trials in line with existing amyloid PET qualification opinions.

At the same time, EMA explicitly kept prognostic and predictive use outside the Context of Use.

That restraint is worth noticing.

QoNM is not expanding by lowering standards. It is expanding by drawing sharper boundaries around what a method can credibly support. 

The strategic read

The temptation is to interpret all of this as a story about openness to innovation. That is too vague to be useful.

A more accurate reading is that EMA is building a governance layer for evidence-generating tools that sit upstream of regulatory decisions.

That matters because many of the tools now entering medicine development are neither medicines nor classic devices, yet they shape trial design, endpoint interpretation, patient selection and evidentiary confidence.

QoNM is where some of those boundaries are now being negotiated in public. 

It is also worth noting what this is not.

QoNM has not turned into a high-volume fast lane.

EMA’s mid-year 2025 report said requests for qualification of novel methodologies fell from eight in the first half of 2024 to six in the first half of 2025.

The route remains selective.

At the same time, EMA has widened practical access for not-for-profit developers: certain entities not engaged in economic activity became eligible for fee reductions from January 1, 2025, and EMA states that these reductions also cover qualification advice. The framework is becoming more strategically relevant without becoming easier in evidentiary terms. 

That leaves one clear conclusion for developers, platform owners, consortia and translational teams. The organisations that will benefit most from QoNM are not the ones that treat it as a late-stage validation badge. They are the ones that define a narrow Context of Use early, separate the real claim from the aspirational one, and build the evidence package around that claim from the start.

The smart move is not to ask whether a novel method is impressive. It is to ask whether it can survive qualification on terms precise enough for regulators to use.

That is where the strategic value now sits.

Executive takeaways

Monitor now. QoNM is becoming a practical signal for how EMA intends to handle AI, modelling, imaging and other evidence tools before they become central to product review.

Do not assume. Novelty does not carry the process. Context of Use, evidentiary discipline and lifecycle governance do.

Who should care. This is not just a pharma issue anymore. It now matters to academic method developers, public-private consortia, data-source owners, medtech-adjacent platform teams and investors watching which evidence infrastructures are becoming regulator-ready.

Where attention belongs. Not on generic claims about innovation, but on the emerging rules for re-validation, scoping, output templates and method-specific guidance.

#EMA #QoNM #QualificationOfNovelMethodologies #RegulatoryScience #DrugDevelopment #BiomedicalAI #PBPK #ClinicalTrials #EvidenceGeneration #MedicalInnovation #MASH #AlzheimersResearch

References

• European Medicines Agency. (2023, December 15). Report on the EMA multi-stakeholder workshop on qualification of novel methodologies. 

• European Medicines Agency. (2023, October 4). Qualification of novel methodologies for drug development: Guidance to applicants (Rev. 5). 

• European Medicines Agency. (2024, September 4). Future-proofing Qualification of Novel Methodologies (QoNM): Action plan. 

• European Medicines Agency. (2024, June 25). Qualification opinion for Centiloid measure of Amyloid PET to quantify brain amyloid deposition. 

• European Medicines Agency. (2025, March 20). EMA qualifies first artificial intelligence tool to diagnose inflammatory liver disease (MASH) in biopsy samples. 

• European Medicines Agency. (2025, August 1). Qualification opinion for Simcyp Simulator. 

• European Medicines Agency. (2025, September 22). Mid-year report 2025: January–June 2025. 

• European Medicines Agency. (2025, December). Future-proofing Qualification of Novel Methodologies (QoNM): Update on delivery of the action plan [Presentation]. 

• European Medicines Agency. (2026, February 2). Highlight report from the 15th Industry Stakeholder Platform on research and development support. 

• European Medicines Agency. (2026, March 20). Qualification opinion for Molecule-independent device bridging approach (MIDBA).