What the EMA reflection paper means for qualifying non-mutagenic impurities

The European Medicines Agency (EMA) has published its final reflection paper on the qualification of non-mutagenic impurities. The guideline clarifies how regulatory, toxicology and CMC teams should approach non-mutagenic impurities (NMIs) in practice, reinforcing the use of structured risk assessment and New Approach Methodologies (NAMs).

This marks a welcome step toward greater clarity in an area that has often relied on expert judgment and case-by-case interpretation.

Non-mutagenic impurities sit in a more nuanced space than mutagenic impurities. While established frameworks exist for mutagenicity, NMIs require broader toxicological evaluation and proportionate reasoning. The new document does not change the underlying science, but it does provide a clearer structure around how teams can build and justify risk-based decisions using existing data, and how to apply new approach methodologies (NAMs).

The reflection paper brings much‑needed structure to impurity assessment, clarifying how NAMs and read‑across should be applied so that decisions are scientifically sound and consistently justified.
Dr Gabriela Silveira, Lead Scientific Consultant

Prioritising NAMs within a structured assessment

A clear message throughout is that teams should consider NAMs as the starting point for qualification. (Q)SAR, read-across, threshold of toxicological concern (TTCs), and in vitro methods are primary tools for building a toxicological justification before in vivo studies are considered.

In impurity assessment, this shift is important. Generating new experimental data for low-level impurities is not always proportionate or practical, so the focus shifts toward making effective use of existing knowledge and computational approaches.

For impurities, it’s not always straightforward to generate new in vitro data, so we expect the computational side of NAMs to grow – along with clearer examples and greater harmonisation in how these approaches are applied.
Dr Martyn Chilton DABT, Scientific Consultant

Building a defensible weight-of-evidence (WoE) package requires appropriate tools and careful expert interpretation of their outputs. The key is not the individual tool, but how the lines of evidence are assembled and interpreted. Structured reasoning, clearly documented, remains central.

When non-mutagenic impurity qualification is required

Qualification becomes necessary when impurity levels exceed the thresholds defined in ICH Q3A or Q3B, taking into account maximum daily dose and exposure considerations. At that point, teams must demonstrate that the impurity is safe at the observed level.

The assessment starts with the ICH Q3A/Q3B qualification thresholds; if an impurity exceeds these, that’s when you need to assess its toxicity to show it’s safe at the level you’re actually seeing.
Dr Martyn Chilton DABT, Scientific Consultant

The reflection paper also provides additional examples of situations that may prompt qualification, including process changes, new degradation pathways or late-emerging impurities. These examples help teams anticipate regulatory expectations earlier in development and lifecycle management.

In practice, effective non-mutagenic impurity qualification combines regulatory thresholds, exposure considerations and structured toxicological assessment into a coherent justification.

In this way, the EMA reflection paper provides a clearer framework for how non-mutagenic impurities should be qualified within pharmaceutical development and regulatory submissions.

Risk assessment in non-mutagenic impurity qualification

Qualification decisions are rarely based on a single parameter. Scientists must consider the level of concern, exposure, clinical context and hazard profile together.

In practice, this means assessing the impurity within the full product context. An impurity that appears low concern based on hazard alone may warrant further consideration when exposure or route of administration is factored in. Equally, exposure may be limited in a way that supports a proportionate conclusion.

The structured flow presented in the guidance supports consistent application of these principles. It provides a framework, but scientific judgement remains central to interpreting the evidence in context.

Interpreting API-like similarity

The discussion around ‘API-like’ impurities is more clearly articulated in the final reflection paper.

API‑likeness cannot rely on structural similarity alone. The key is understanding whether any change influences toxicological behaviour, and the paper provides clearer guidance on how to make that distinction.
Dr Gabriela Silveira, Lead Scientific Consultant

Read-across arguments benefit from this clarification. Scientists should justify similarity in biological and toxicological terms, not solely structural ones.

Applying the metabolite route in impurity qualification

The metabolite pathway remains an important route within non-mutagenic impurity qualification. The guidance clarifies that metabolite status alone is not sufficient and that relative exposure and overall toxicological context should also be considered.

Where an impurity can be demonstrated to be a significant metabolite of the API, this may support qualification. However, this is not automatic. The assessment still needs to consider relative exposure and whether the additional exposure from the impurity meaningfully alters the overall safety profile.

The key is to demonstrate that, considering the patient’s actual exposure, the toxicological profile remains appropriate, always in a proportional and well-justified manner.
Dr Arianne Rivellis, Director of Global Operations

Across each pathway, teams must support their conclusions with transparent reasoning and available data.

Aligning toxicology with manufacturing reality

A particularly practical clarification relates to how impurity specification limits should be justified. Justification should be grounded in batch data and stability observations, rather than theoretical limits alone.

In application, this requires alignment between toxicology, analytical and CMC functions. Qualification decisions should reflect the impurity levels that are observed and the consistency of manufacturing processes.

The strongest justification is always the one grounded in our own data. Basing the qualification on the impurity levels observed in batches and stability studies makes the assessment far more robust and realistic.
Dr Arianne Rivellis, Director of Global Operations

Route of administration shapes the question

Risk cannot be evaluated independently of exposure route. Oral, topical, inhaled and parenteral products present different considerations.

For topical products, sensitisation may be a key endpoint depending on exposure profile and product use. Ensuring that the assessment reflects route-specific considerations supports proportionate and scientifically appropriate conclusions.

Route of administration changes what types of toxicity needs to be considered. With topical products, local effects matter more, and sensitisation is often one of the most sensitive endpoints to consider.
Dr Martyn Chilton DABT, Scientific Consultant

What this means for non-mutagenic impurity qualification

When applying the guidance, five practical considerations remain consistent:

Define clearly what triggers qualification for your product.
Prioritise NAMs and structured weight of evidence approaches.
Ensure endpoints align with route of administration and exposure.
Anchor justifications to observed data from stability and manufacturing.
Document reasoning transparently and proportionately.

We already have a significant body of chemical and toxicological knowledge. The opportunity is to use it better, share information where appropriate, and avoid repeating work unnecessarily.
Dr Arianne Rivellis, Director of Global Operations

The principles reinforced here are consistent with the approach Consult Lhasa applies when supporting impurity qualification.

Our teams of expert scientific consultants work with industry partners to build structured, transparent justifications for non-mutagenic impurities, combining in silico tools, read-across strategies and expert toxicological evaluation to ensure decisions are proportionate and clearly documented.