The control and qualification of impurities remain among the most critical aspects in the development and lifecycle management of pharmaceutical products. However, the recently published reflection paper from the European Medicines Agency signals an important evolution in how non-genotoxic impurities should be assessed, justified, and managed during pharmaceutical development.
Until recently, qualification of non-genotoxic impurities frequently relied on additional animal studies, often providing limited impurity specific information. The new EMA framework instead promotes a more integrated and science-driven strategy based on risk assessment, computational toxicology, read across approaches, and New Approach Methodologies (NAMs).
For pharmaceutical companies, this represents both an opportunity and a challenge: reducing unnecessary testing while simultaneously increasing the need for robust toxicological expertise and transparent scientific justification.
What Are Non-Genotoxic Impurities (NGI)?
Non-genotoxic impurities are chemical entities that are not expected to directly interact with DNA or induce gene mutations. Unlike mutagenic impurities regulated under International Council for Harmonisation, these compounds are primarily addressed under the principles of ICH Q3A(R2) and ICH Q3B(R2).
Despite lacking genotoxic potential, non-genotoxic impurities may still present toxicological concerns depending on: pharmacological activity, systemic toxicity, exposure duration, bioaccumulation potential, physicochemical properties, target organ toxicity.
Such impurities may originate from: starting materials, intermediates, degradation products, catalysts and reagents, synthetic by products, residual solvents, process-related contaminants, extractables and leachables
In recent years, increasing manufacturing complexity, particularly in HPAPIs, synthetic peptides, and novel pharmaceutical technologies, has made impurity qualification substantially more challenging.
The recent publication of the reflection paper by the European Medicines Agency provides additional clarity on how non-mutagenic impurities should be evaluated and justified during pharmaceutical development and regulatory submissions. Although the scientific principles underlying impurity assessment remain unchanged, the document introduces a more structured framework for decision making, with a stronger emphasis on transparent risk assessment and the appropriate use of modern toxicological tools.
Why Is the EMA Reflection Paper Important?
The EMA reflection paper provides a more structured framework for evaluating non-genotoxic impurities, especially when impurities exceed ICH qualification thresholds or are identified after completion of non-clinical studies.
Importantly, the document does not replace existing ICH guidelines. Instead, it complements them by introducing a modern and risk-based interpretation of impurity qualification.
One of the most important messages emerging from the paper is that impurity qualification should no longer be viewed as a purely threshold driven exercise. Instead, the EMA emphasizes integrated toxicological assessment based on the totality of scientific evidence.
The document strongly promotes: Weight of Evidence approaches, computational toxicology, read across methodologies, metabolite qualification strategies, threshold-based concepts, in vitro approaches, exposure driven assessment
At the same time, the EMA clearly stresses that these tools require expert interpretation and cannot be applied in a purely mechanistic or automated manner.
How Does the EMA Reflection Paper Structure the Qualification Process?
The reflection paper proposes a stepwise and risk-based workflow for impurity qualification.
The approach becomes relevant when:
- impurity levels exceed ICH Q3A(R2) or Q3B(R2) qualification thresholds
- existing non clinical studies do not adequately cover the impurity
- new impurities emerge during manufacturing changes or stability studies
A key concept introduced by the EMA is that not all impurities require the same level of toxicological investigation. For example, impurities may already be considered qualified when:
- they are also observed as metabolites in humans or animals at sufficient exposure levels
- they are structurally similar to the API and do not introduce new toxicophores
- they are considered “API-like” and toxicologically covered by existing drug substance studies
This is particularly relevant for oxidation or hydrolysis products that retain most of the parent structure without significantly altering toxicological behaviour.
Conversely, impurities presenting relevant structural differences, altered pharmacokinetic behaviour, or new toxicophores may require additional toxicological assessment.
The reflection paper also emphasizes that qualification should always consider the broader clinical context, including: route of administration, patient population, duration of treatment, expected systemic exposure, and severity of disease
This represents a significant shift toward product specific and exposure-based toxicological evaluation.
What Does the EMA Reflection Paper Add to Existing Impurity Guidelines?
The reflection paper proposes a stepwise and risk-based strategy for the qualification of non-genotoxic impurities. The approach applies when impurity levels exceed the qualification thresholds defined in ICH Q3A(R2) or ICH Q3B(R2), and when existing non-clinical studies with the drug substance do not adequately cover the impurity.
A central concept of the document is that not all impurities require the same level of toxicological investigation. For example, impurities that are also observed as metabolites in animals or humans at sufficient exposure levels may be considered already qualified without the need for additional toxicological studies. Similarly, API related impurities that do not introduce new structural alerts or toxicophores compared to the parent compound are generally considered of low toxicological concern and potentially covered by existing studies performed with the drug substance.
For impurities that cannot be addressed through these simplified approaches, the reflection paper recommends a broader evaluation of the level of toxicological concern. This assessment should consider factors such as patient population, route of administration, duration of treatment, and expected exposure. If the overall assessment indicates low concern, no further qualification may be necessary and the impurity can be considered toxicologically acceptable.
When impurity-specific safety information is required, alternative strategies to gather this information may be applied, including the use of Threshold of Toxicological Concern (TTC), in silico methods ((Q)SAR), read-across (RAX) and in vitro approaches. A Weight of Evidence (WoE) approach including an assessment of the level of uncertainty may be used to decide whether the NGI can be considered safe at the specified level.
It should be noted that impurities that are structurally similar to the API and retain most of its functional groups can be considered API-like, such as clobetasol propionate, clobetasone-17-propionate, and betamethasone. This includes degradation products from oxidation or hydrolysis that introduce minor modifications without significantly altering the overall structure or size of the molecule. Additionally, dimer or trimer impurities may be considered API-like if the dimerization bridge does not introduce a new toxicophore and the impurity degrades back into the parent compound in systemic circulation (e.g., empagliflozin sugar dimer). If degradation is unlikely, further qualification of the impurity should assess its altered physicochemical properties and biological activity.
To evaluate similarity, computational predictive tools can identify toxicophores and predict physicochemical (PC) and pharmacokinetic (PK) properties, comparing the impurity to the API. If no new toxicophores are introduced and PC/PK parameters remain unchanged, the impurity is considered API-like, meaning its toxicological properties are covered by existing non-clinical studies, and no further investigations are needed.
In general, API-derived molecules without new structural alerts are considered of low toxicological concern and they are considered covered by the toxicological studies of the API.
Conversely, if impurities show significant structural differences or alterations in PC/PK properties that could impact toxicity, or if computational tools predict such differences, the impurity should be classified as non-API-like. Exceptions exist where small structural changes significantly alter toxicity, creating activity cliffs. This is particularly relevant when toxicity depends on binding to a specific site.
What Happens When Additional Qualification of an Impurity Is Required?
When the initial assessment indicates that a non-genotoxic impurity may require further evaluation, the EMA reflection paper recommends a stepwise and science driven strategy aimed at minimizing unnecessary animal testing.
The first step should always involve the collection of existing information. This includes a literature review for available impurity specific in vitro or in vivo toxicological data that may support qualification of the impurity at the proposed exposure level.
If adequate in vivo data are available, an Acceptable Level (AL) can be derived using toxicological principles similar to those used for PDE calculations. When the impurity level remains below the derived AL, the impurity may be considered toxicologically qualified.
For data-poor substances, the reflection paper encourages the use of surrogate compounds through scientifically justified read-across approaches. In this context, toxicological data generated for structurally related substances may support derivation of an AL for the impurity.
If neither direct data nor surrogate information are available, the document strongly promotes the use of New Approach Methodologies (NAMs), including:
- computational toxicology tools such as QSAR models
- in vitro toxicological approaches
- integrated Weight of Evidence assessments
- Adverse Outcome Pathway (AOP) based evaluations
The reflection paper also emphasizes that NAM outputs should always be interpreted within a broader toxicological context, taking into account uncertainty, predicted hazard, exposure level, and product specific considerations.
Importantly, the EMA clearly states that additional in vivo studies should only be considered as a last resort. According to the proposed framework, animal studies may become appropriate only in exceptional situations where all alternative approaches fail to provide sufficient information to adequately qualify the impurity.
What Does This Mean for Pharmaceutical Companies?
From a regulatory perspective, the EMA reflection paper represents a major evolution toward modern, integrated, and scientifically justified impurity assessment.
For pharmaceutical companies, this creates several important implications.
On one hand, the new framework may reduce the need for unnecessary animal testing and provide greater flexibility in managing: post approval changes, manufacturing modifications, degradation products, complex impurity profiles, peptide and HPAPI development.
On the other hand, the reflection paper significantly increases the importance of:
- computational toxicology expertise
- scientifically robust read across strategies
- integrated Weight of Evidence approaches
- transparent uncertainty analysis
- exposure driven toxicological assessment
In practice, successful impurity qualification will increasingly require strong integration between chemistry, toxicology, pharmacokinetics, and regulatory strategy.
Final Considerations
The EMA reflection paper clearly signals the future direction of pharmaceutical impurity qualification: less reliance on default animal testing, greater integration of NAMs, and increasingly sophisticated risk-based toxicological assessment.
Rather than introducing rigid new testing requirements, the document encourages scientifically justified, product-specific, and exposure-driven evaluation strategies supported by the totality of available evidence.
For companies developing complex pharmaceutical products, this evolution offers important opportunities to optimize impurity qualification strategies while maintaining robust patient safety standards.
At ToxHub, these approaches are routinely integrated into toxicological risk assessment strategies supporting pharmaceutical development, HPAPI qualification, read across justifications, computational toxicology assessment, and regulatory impurity evaluations.
Issued by Chiara Gazerro