Expertise · Clinical & Preclinical Studies

Preclinical Studies for Biological and Biosimilar Medicinal Products in the EAEU

Development programmes supporting medicinal product registration

This section brings together clinical, bioequivalence and preclinical evidence used to support medicinal product registration, including local-study strategy, foreign-data usability, GCP/GLP inspection risk, biowaivers and biological products.

Detailed guidance across the evidence programme is available in the current technical library.

Clinical and preclinical evidence strategy supporting EAEU registration

Biological medicinal products should not be forced into a small-molecule toxicology template.

A monoclonal antibody, recombinant therapeutic protein, vaccine, plasma-derived product, low-molecular-weight heparin, cell-based therapy or other complex biological product can raise nonclinical questions that simply do not exist for a conventional tablet such as paracetamol.

The EAEU therefore regulates biological development through both the general registration/GLP framework and the product-specific Rules for Studies of Biological Medicinal Products under Decision No. 89.

As with conventional pharmaceuticals, there is no general requirement that the preclinical studies be conducted in Russia or another EAEU member state.

The requirement is that the Module 4 evidence—together with the biological comparability and quality package—meets the applicable EAEU requirements and supports the proposed benefit-risk conclusion.

For a biosimilar, the central nonclinical question is especially different:

What residual uncertainty remains after analytical and functional comparability, and does that uncertainty actually require an in-vivo animal study?

That question should be answered before a conventional repeat-dose toxicity programme is ordered by default.

Biological development starts in Module 3, not Module 4

For a small molecule, a chemical structure can often define the active substance with high precision.

For a biological product, the molecule is inseparable from its production system and higher-order characteristics.

Relevant attributes can include:

  • primary structure;
  • higher-order structure;
  • glycosylation pattern;
  • charge variants;
  • aggregation;
  • biological activity/potency;
  • receptor binding;
  • Fc-mediated functions where relevant;
  • process-related impurities;
  • product-related variants;
  • formulation and container interactions.

For a biosimilar, extensive analytical and functional comparability is therefore the foundation of the development programme.

The nonclinical programme should address what remains uncertain after that comparison.

A weak analytical comparability package cannot be compensated for by a large animal study.

Decision No. 89 creates product-specific development logic

Decision No. 89 establishes the EAEU rules for studying biological medicinal products.

The framework has been expanded over time. Amendments adopted in 2022 and 2023 added or refined requirements for additional biological groups, including plasma-derived products, therapeutic proteins, vaccines and related specialized safety issues. In January 2025 the EEC added two further chapters covering high-technology products based on genetically modified somatic cells, including advanced anti-tumor cell products.

This continuing expansion matters because “biological medicinal product” is not one homogeneous regulatory class.

A biosimilar monoclonal antibody, prophylactic vaccine and genetically modified cell therapy cannot share one standard nonclinical checklist.

The programme must follow the product-specific chapter and mechanism.

Biosimilar development is comparative from the beginning

A biosimilar is not developed as an independent new biological molecule whose safety and efficacy must be established from zero.

The regulatory objective is to demonstrate a high degree of similarity to an appropriate reference biological medicinal product and show that remaining differences have no clinically meaningful effect on quality, safety or efficacy.

The evidence hierarchy is therefore:

analytical comparability → functional comparability → nonclinical assessment → clinical PK/PD → comparative clinical/immunogenicity evidence where required.

The better the early comparability evidence, the smaller the residual uncertainty that the later programme has to address.

This is the opposite of a “do every study in parallel” model.

In-vitro functional comparison is often more informative than an animal model

Biological activity can frequently be compared with high sensitivity through product-specific in-vitro systems.

Depending on the molecule, these can include:

  • target/receptor binding;
  • ligand neutralization;
  • enzyme activity;
  • cell proliferation/inhibition;
  • apoptosis/cytotoxicity;
  • Fc receptor binding;
  • antibody-dependent cellular cytotoxicity;
  • complement-dependent cytotoxicity;
  • other mechanism-specific functional assays.

For a biosimilar, a panel of orthogonal functional assays can detect meaningful differences more directly than an animal species whose target biology is only partially relevant.

Where analytical and in-vitro functional comparability are convincing, the need for an in-vivo study can be reduced.

The RegLek-2025 benefit-risk session specifically addressed current EAEU requirements for in-vivo preclinical studies of different classes of biosimilar medicinal products, reflecting the regulator’s current focus on when animal work is and is not scientifically justified.

An in-vivo biosimilar study should answer a residual question

Animal studies for biosimilars are not intended to reproduce the full development programme of the reference biological product.

An in-vivo study can be justified where:

  • the mechanism cannot be characterized adequately in vitro;
  • analytical/functional comparison identifies residual uncertainty;
  • there is a product-class-specific EAEU expectation;
  • exposure or toxicity cannot be interpreted from existing data;
  • the formulation or impurity profile creates a new safety concern;
  • there is a relevant species/model capable of providing meaningful information.

If none of these applies, an animal study may add little regulatory value.

The study should therefore be designed as comparative and hypothesis-driven, not as a generic toxicology package copied from an original biologic.

Relevant species selection is critical for biological products

Species selection is often straightforward for small molecules and much more difficult for biologics.

A biologic may bind only to human or primate target, may have different receptor affinity across species or may lack pharmacological activity altogether in conventional laboratory animals.

A species is relevant when the test product produces a meaningful pharmacological response through the mechanism of interest.

Selection should consider:

  • target expression;
  • receptor/ligand homology;
  • binding affinity;
  • downstream functional response;
  • tissue distribution;
  • PK/exposure;
  • known class biology.

Testing in a non-relevant species simply to produce a second-species report can generate misleading data and unnecessary animal use.

Where no conventional species is relevant, alternative models, homologous molecules or other scientifically justified strategies may be needed depending on the product.

The regulatory justification is as important as the model itself.

Repeat-dose toxicity can integrate several endpoints

For many biologics, where an in-vivo study is justified, a well-designed repeated-dose toxicity study can incorporate multiple objectives.

The programme can include:

  • clinical observations;
  • clinical pathology;
  • histopathology;
  • toxicokinetics;
  • pharmacodynamic markers;
  • local tolerance;
  • immune-response/anti-drug antibody assessment;
  • recovery/reversibility where relevant.

For a biosimilar, the design should generally be comparative against the reference biological product where the study is intended to address biosimilarity-related uncertainty.

The focus is not merely whether either product causes toxicity. It is whether any difference between the products is biologically meaningful.

Animal immunogenicity has a specific but limited role

Biological products can induce anti-drug antibodies (ADA) in animals.

That information can be important in a nonclinical study because ADA may:

  • change systemic exposure;
  • neutralize biological activity;
  • alter pharmacodynamic response;
  • create apparent differences in toxicity;
  • make later study time points difficult to interpret.

For that reason, immunogenicity/ADA assessment can be integrated into comparative repeated-dose studies where relevant.

However, animal immunogenicity is generally not predictive of human immunogenicity.

A lower or higher ADA incidence in animals should not be presented as proof that the biosimilar will be less or more immunogenic in patients.

Comparative human immunogenicity remains a clinical-development issue where required by the applicable EAEU biosimilar framework.

The value of the animal ADA data is primarily to interpret the nonclinical PK/PD and toxicity results.

Comparative immunogenicity has to be understood at two levels

For a biological medicinal product, the phrase “comparative immunogenicity” can refer to two different regulatory questions.

Preclinical level

Does immune response in the animal model alter exposure or interpretation of comparative toxicity/PD data?

This is assessed through appropriate ADA sampling and interpretation where scientifically relevant.

Clinical level

Do the biosimilar and reference product show clinically comparable immunogenicity in humans, including binding and, where relevant, neutralizing antibodies and their impact on PK, efficacy and safety?

This belongs to the clinical comparability programme.

Conflating the two creates a common dossier error: overinterpreting animal ADA results as evidence of clinical similarity.

Pharmacology and tissue cross-reactivity can be product-specific

For monoclonal antibodies and some other biological products, tissue cross-reactivity or target-distribution studies can be relevant to species selection and safety interpretation.

The value depends on product mechanism, target expression, available knowledge and the design of the overall programme.

Similarly, safety pharmacology may be integrated into general toxicity studies where that is scientifically appropriate rather than performed as a completely separate battery.

The product-specific EAEU framework should drive the choice.

Genotoxicity and carcinogenicity are often different for biologics

Conventional genotoxicity assays designed for small chemical molecules are often not scientifically appropriate for large biological molecules such as therapeutic proteins because they do not interact with DNA through the same mechanisms and are degraded to peptides/amino acids.

Similarly, conventional lifetime rodent carcinogenicity studies may be inappropriate for many biologics.

That does not mean carcinogenicity risk is irrelevant.

For a biologic affecting growth factors, immune regulation or proliferative pathways, tumor risk may need to be assessed through mechanism, class data, pharmacology, chronic toxicity, clinical data or other product-specific approaches.

The regulatory principle is scientific relevance, not automatic exemption.

Reproductive and developmental toxicity requires relevant pharmacology

For a biologic intended for women of childbearing potential or chronic use, reproductive/developmental risk may need to be characterized.

Species selection again matters because target pharmacology may be species-specific.

Study design should account for:

  • placental transfer where relevant;
  • target expression during development;
  • maternal pharmacology;
  • exposure in offspring;
  • neonatal immune-development considerations for immune-modulating products;
  • class effects.

For some monoclonal antibodies, placental transfer changes markedly with gestational stage, which can influence study design and clinical risk interpretation.

The final purpose is to support the pregnancy/lactation and fertility sections of the product information.

Vaccines use immunogenicity as a pharmacological endpoint

For vaccines, nonclinical immunogenicity has a different meaning from ADA assessment against a therapeutic protein.

The programme can assess whether the vaccine induces the intended immune response and can include:

  • antibody responses;
  • neutralizing activity;
  • cellular immune responses;
  • dose-response;
  • schedule effects;
  • adjuvant contribution;
  • challenge/protection models where relevant and available.

Safety assessment also has to consider the vaccine platform, adjuvants, route, repeat dosing and intended population.

A vaccine is therefore a good example of why a biological Module 4 cannot be reduced to a conventional toxicity checklist.

Plasma-derived and other biological classes have additional safety questions

Products derived from human plasma and other biological sources raise specific questions beyond ordinary systemic toxicology.

The EAEU biological rules include dedicated requirements for issues such as viral safety and, for relevant product classes, reduction of transmissible-agent risks.

The 2023 amendments to Decision No. 89 expanded specific rules for plasma-derived and biotechnology-derived products and introduced measures addressing prion-transmission risk for relevant groups.

These safety issues sit at the boundary between Module 3 and Module 4 and must be handled as an integrated biological safety strategy.

Cell and gene-related products now have more specific EAEU rules

Decision No. 13 of 22 January 2025 supplemented Decision No. 89 with new chapters for high-technology medicinal products based on genetically modified somatic cells.

For such products, classical animal-toxicology concepts are often insufficient.

The development programme can require product-specific assessment of:

  • cell phenotype/function;
  • biodistribution;
  • persistence;
  • ectopic tissue localization;
  • tumorigenicity/oncogenic risk;
  • off-target effects;
  • vector-related risks;
  • immunological effects;
  • dose and route relevance.

The regulatory field is evolving quickly, so a programme designed several years ago should be rechecked against the current version of Decision No. 89 before EAEU filing.

For an original biologic, the programme is broader than for a biosimilar

The distinction between an original biological medicinal product and a biosimilar must remain explicit.

An original biologic has no established reference-product safety/efficacy bridge that can replace its development programme. It requires a product-specific package supporting first-in-human development and final registration.

A biosimilar, by contrast, leverages the established clinical experience of the reference biological product and focuses on demonstrating similarity.

Therefore:

original biologic → characterize product-specific hazards and support the full development programme;

biosimilar → identify residual uncertainty after comparability and test only what is needed to resolve it.

Using the original-biologic template for a biosimilar can generate unnecessary animal studies and obscure the comparative logic expected by regulators.

The test material must represent the comparability state

For biosimilars especially, the nonclinical test and reference materials should be linked to the analytical comparability programme.

The dossier should document:

  • test/reference batch identity;
  • quality attributes;
  • potency/biological activity;
  • impurity profile;
  • formulation;
  • stability;
  • manufacturing stage;
  • relation to clinical and commercial batches.

If the biosimilar manufacturing process changes substantially after the pivotal nonclinical or clinical comparability studies, a new comparability assessment may be required.

The answer is not automatically to repeat every study. It is to determine whether the process change creates a new difference relevant to safety, efficacy or immunogenicity.

Module 4 should not duplicate Module 3 comparability

A good biosimilar dossier separates but connects the evidence streams.

Module 3 demonstrates structural, physicochemical and biological similarity.

Module 4 addresses nonclinical pharmacology/safety uncertainty that remains relevant after Module 3.

Module 5 then addresses clinical PK/PD, efficacy/safety and immunogenicity questions that remain after Modules 3 and 4.

The development logic should therefore read as one sequence:

what we know from quality → what uncertainty remains → what the nonclinical programme resolves → what still needs clinical confirmation.

A dossier that repeats the same descriptive comparison in every module without showing this reduction of uncertainty looks larger but not stronger.

Foreign biological nonclinical studies are acceptable if they meet EAEU requirements

There is no general territorial requirement to repeat biological preclinical studies inside the EAEU.

Foreign studies can be used when they are scientifically appropriate, GLP-compliant where applicable and sufficiently documented for expert review.

The sponsor should be able to demonstrate:

  • relevant EAEU product-class requirements were addressed;
  • the species/model was pharmacologically relevant;
  • comparative design was appropriate for a biosimilar;
  • test and reference materials were characterized;
  • GLP status is clear;
  • the laboratory/test sites can support inspection if required;
  • Module 4 conclusions are consistent with Modules 3 and 5.

The geography of the laboratory is not the problem. Regulatory usability and inspectability are.

GLP inspections apply to biological Module 4 studies too

The current EAEU GLP inspection framework under Decisions No. 83 and No. 60 applies to safety studies included in Module 4 irrespective of whether the product is a small molecule or biologic.

NCESMP states that preclinical pharmaceutical inspections:

  • are appointed only on an unscheduled basis during the regulatory procedure;
  • are risk-based;
  • select one of the safety studies included in Module 4;
  • can involve one or several test facilities/sites depending on how the study was outsourced;
  • can combine on-site and documentary inspection subjects;
  • do not suspend dossier expertise for the duration of the unscheduled inspection.

For a biosimilar or other biologic, an inspection can therefore focus on a pivotal comparative toxicity/safety study if its reliability becomes material to the benefit-risk assessment.

Biological studies have additional inspection vulnerabilities

In addition to standard GLP issues, complex biological studies can raise traceability questions around:

  • biological activity of test/reference batches;
  • assay validation;
  • ADA assays;
  • toxicokinetic assays;
  • neutralizing-antibody methods where used;
  • target/receptor characterization;
  • cell-based potency assays;
  • handling of biological reference materials;
  • sample stability;
  • batch comparability;
  • pathology interpretation in species with target-mediated findings.

A study can be well conducted operationally but become difficult to interpret if the bioanalytical or immunogenicity methods were not fit for purpose.

Pre-inspection readiness should therefore include the specialized assays, not only animal-room GLP documentation.

Critical findings can invalidate a pivotal safety bridge

NCESMP’s current inspection guidance states that critical findings in an unscheduled GLP inspection—or an unsatisfactory CAPA response—can prevent a positive registration, renewal or variation decision.

For a biosimilar, the impact can be disproportionate if the inspected study is the principal nonclinical evidence used to resolve a residual comparability concern.

The manufacturer should therefore assess whether the pivotal study remains:

scientifically interpretable + GLP traceable + vendor traceable + analytically traceable + archivally retrievable.

If one of those layers is weak, the risk should be addressed before submission.

Where biological preclinical programmes most often go wrong

Common problems include:

  • treating a biosimilar like an original biologic and commissioning a full new toxicology programme;
  • treating a biologic like a small molecule and selecting non-relevant species;
  • performing animal studies before analytical/functional comparability is mature;
  • using an in-vivo study that does not address a defined residual uncertainty;
  • interpreting animal ADA incidence as proof of comparative human immunogenicity;
  • selecting a species based on availability rather than target pharmacology;
  • ignoring product-class-specific chapters of Decision No. 89;
  • failing to integrate test/reference batch characterization with Module 3;
  • allowing manufacturing changes to break the link between nonclinical, clinical and commercial material;
  • using legacy GLP reports whose specialized bioanalytical/ADA records are no longer inspectable;
  • assuming a foreign nonclinical study must be repeated solely because it was not conducted in the EAEU.

The strongest biological programme is usually the one that removes unnecessary animal studies while making each remaining study scientifically decisive.

What Pharegis manages for biological and biosimilar Module 4 programmes

Typical deliverables include:

  • biological-product classification and Decision No. 89 requirements map;
  • original-biologic vs biosimilar development-pathway assessment;
  • reference biological product strategy for biosimilars;
  • Module 3 → Module 4 residual-uncertainty map;
  • assessment of whether in-vivo comparative studies are actually required;
  • relevant-species/model justification;
  • comparative pharmacology strategy;
  • repeat-dose toxicity / toxicokinetic design review;
  • ADA/immunogenicity assay and interpretation strategy for nonclinical studies;
  • distinction between animal ADA findings and human comparative immunogenicity requirements;
  • reproductive/developmental or other product-specific safety strategy;
  • vaccine, plasma-derived, protein or advanced-therapy product-specific gap assessment;
  • foreign GLP-study usability and inspectability review;
  • test/reference batch and CMC comparability assessment;
  • integrated Module 2/4 narrative;
  • pre-GLP-inspection readiness review;
  • support during NCESMP unscheduled GLP inspection and CAPA;
  • expert-query response strategy.

For a biosimilar manufacturer, the objective is not to prove safety from zero. It is to demonstrate that the biosimilar’s remaining uncertainty has been reduced enough that the reference product’s established clinical knowledge can be relied upon.

Discuss a biological / biosimilar preclinical programme

For an initial assessment, prepare:

  • biological product class and mechanism;
  • original vs biosimilar status;
  • proposed reference biological product;
  • analytical/functional comparability summary;
  • Module 4 study list and reports;
  • relevant-species justification;
  • toxicokinetic/PD data;
  • ADA/immunogenicity data;
  • GLP status and laboratories/test sites;
  • test/reference batch characterization;
  • planned or completed clinical comparability programme;
  • manufacturing/process changes since pivotal studies;
  • target EAEU filing date.

Pharegis can then determine which EAEU product-specific requirements apply, whether additional animal studies are scientifically justified, whether the existing foreign evidence is usable and whether the pivotal Module 4 studies present a GLP-inspection risk.

The objective is not to repeat biological development in the EAEU. It is to build a product-specific evidence chain in which quality comparability, nonclinical evidence and clinical confirmation reduce uncertainty in the correct order.

Regulatory basis

Regulatory status reviewed: August 2026.