Biosimilars in Europe: Comparative Evidence and Regulatory Strategy

The European Union’s regulatory architecture for biosimilars represents one of the most mature and scientifically rigorous frameworks globally, established to foster competition while safeguarding patient safety. Originating from the overarching pharmaceutical legislation and refined through the European Medicines Agency’s (EMA) extensive scientific guidance, the framework is predicated on a fundamental principle: demonstrating highly similar quality, non-clinical, and clinical characteristics to a reference medicinal product, rather than re-establishing the full totality of evidence required for a novel biologic. This approach acknowledges the inherent variability and complexity of biotechnology-derived medicines, where absolute identity is unattainable. Consequently, the regulatory strategy focuses on a stepwise, evidence-based assessment of residual uncertainties, culminating in a positive benefit-risk assessment for marketing authorisation.

The Legal and Scientific Foundation of Biosimilarity

The legal basis for biosimilar development in the EU is anchored in Directive 2001/83/EC and Regulation (EC) No 726/2004. A biosimilar is defined as a medicinal product that contains a version of the active substance of an already authorised original biological medicinal product (reference product). Crucially, the EMA does not employ the term ‘interchangeable’ in the same manner as the US FDA; instead, it relies on the concept of ‘interchangeability’ being a matter for national competent authorities (NCAs) and prescribing physicians, based on the totality of evidence submitted. The core scientific guidance is provided by the EMA’s overarching ‘Guideline on similar biological medicinal products containing biotechnology-derived proteins as active substance’ and specific product-class guidelines.

From a practitioner’s perspective, the regulatory pathway is not a linear checklist but a holistic evaluation. The EMA prioritises a comparative exercise that begins with the most sensitive analytical methods. The agency explicitly states that the more extensive the structural and functional characterisation, the lower the risk of clinically meaningful differences. This represents a shift from the traditional drug development paradigm where clinical efficacy trials are the primary source of evidence; in biosimilars, analytical similarity is the bedrock upon which clinical data rests.

The Stepwise Approach to Demonstrating Comparability

The EMA mandates a stepwise approach to minimise the residual risk of immunogenicity or efficacy differences. This approach is designed to be scientifically robust and resource-efficient.

Step 1: Quality (Physicochemical and Biological Characterisation)

This is the most critical phase. Developers must compare the biosimilar candidate to the reference product using a battery of state-of-the-art analytical techniques. This includes assessing the primary amino acid sequence, post-translational modifications (such as glycosylation patterns), higher-order structures, and biological activities (e.g., receptor binding, cell-based potency assays). The goal is to demonstrate that the product is highly similar to the reference product, identifying and justifying any observed differences. If the analytical data reveal significant disparities, the development program is likely to halt here.

Step 2: Non-Clinical Studies

While the EMA has moved towards waiving or reducing animal studies, non-clinical in vivo studies (toxicity testing) may still be required if a specific risk is identified during the analytical assessment. However, the guideline emphasises that in vitro pharmacodynamic assays and physicochemical characterisation often provide more relevant information than animal testing for large molecules. This reflects the EU’s commitment to the 3Rs (Replacement, Reduction, and Refinement) principles.

Step 3: Clinical Studies

Even if analytical and non-clinical data show high similarity, clinical studies are almost always required to detect potential subtle differences that may impact safety and efficacy, particularly regarding immunogenicity. The clinical package typically includes:

• Pharmacokinetics (PK) and Pharmacodynamics (PD): A comparative study in a sensitive population (often healthy volunteers) to demonstrate equivalent exposure and biological effect.
• Efficacy and Safety (Immunogenicity): A comparative clinical trial in a sensitive patient population to confirm no clinically meaningful differences in efficacy and to evaluate the immunogenicity profile.

The design of these trials is highly specific. For instance, the EMA often requires a parallel group design rather than a crossover design for immunogenicity assessment, as this provides a clearer signal of antibody formation over time. Furthermore, the choice of patient population is vital; it must be sensitive enough to detect differences but ethically justifiable.

Comparative Evidence: The Heart of the Regulatory Submission

The concept of ‘comparative evidence’ is the linchpin of the EU biosimilar pathway. It is not enough to show that the biosimilar meets pre-specified quality standards; it must be proven to be comparable to the reference product. This comparative nature permeates every stage of development.

In the analytical phase, this involves side-by-side testing of the proposed biosimilar and the reference product, often using multiple lots of the reference product to account for its natural batch-to-batch variability. The EMA looks for the similarity range—the variability observed with the reference product—to serve as the benchmark against which the biosimilar’s variability is judged. If the biosimilar falls within this range, it is considered highly similar.

In the clinical phase, the comparative evidence is statistical. The primary endpoints for PK/PD studies must meet strict equivalence criteria (typically a 90% confidence interval within 80-125%). For efficacy trials, the EMA generally accepts a pre-specified margin for the primary efficacy endpoint that is clinically relevant. However, the agency places equal weight on safety endpoints, particularly the incidence of anti-drug antibodies (ADAs). A higher rate of immunogenicity in the biosimilar arm compared to the reference arm is a major red flag that requires extensive investigation.

Extrapolation of Indications

One of the most strategic advantages of the EU pathway is the potential for extrapolation. If a biosimilar is approved for one indication of the reference product, it may be approved for other indications without conducting clinical trials in those specific populations, provided that:

1. The mechanism of action is the same across indications.
2. The safety and efficacy profile is expected to be consistent.
3. The analytical and clinical data support the assumption of similarity for the extrapolated indications.

For example, a biosimilar to a monoclonal antibody approved for rheumatoid arthritis might be extrapolated to oncology indications if the receptor binding and functional activity (e.g., ADCC) are confirmed in the analytical package. This requires a robust scientific justification in the marketing authorisation application (MAA).

Clinical Requirements: Beyond Efficacy to Immunogenicity

While efficacy is a secondary concern in biosimilar development (the assumption is that if it is highly similar, it will be equally efficacious), safety—specifically immunogenicity—is the primary driver of clinical requirements. Biologics can elicit immune responses that neutralise the drug or cause adverse reactions. The EMA’s clinical requirements are designed to detect any increase in immunogenicity risk.

The standard clinical trial design involves a randomised, double-blind, parallel-group study. The duration of the study is often determined by the known immunogenicity profile of the reference product. For monoclonal antibodies, a duration of at least 52 weeks is often required to capture the kinetics of antibody formation.

Furthermore, the EMA places significant emphasis on the state of the art assays for detecting ADAs. Developers must validate their assays to ensure they are sensitive and drug-tolerant. The interpretation of immunogenicity data is nuanced; the agency looks not only at the incidence of ADAs but also at their neutralising potential and clinical impact (e.g., loss of efficacy or hypersensitivity reactions).

Switching and Interchangeability: The National Nuance

The distinction between ‘biosimilar’, ‘switching’, and ‘interchangeability’ is a source of frequent confusion and regulatory divergence within Europe. It is vital to distinguish between these concepts.

Switching refers to the decision by a treating physician to change a patient’s treatment from one biologic to another (e.g., from the reference product to a biosimilar, or between two biosimilars). This is a medical decision.

Interchangeability refers to the administrative or automatic substitution of one medicine for another without the intervention of the prescriber. In the EU, the EMA does not assess or grant an ‘interchangeable’ status. This is a competence retained by Member States.

However, the EMA has issued a Statement on the interchangeability of biosimilar medicines, which provides a scientific opinion that supports the use of biosimilars as interchangeable in clinical practice. The statement highlights that the rigorous comparability exercise ensures that there is no clinically relevant difference in safety and efficacy between the biosimilar and the reference product.

Despite this EU-level scientific consensus, national implementation varies significantly:

• Germany: Has a system of ‘automatic substitution’ where pharmacists can substitute a biosimilar for a prescribed biologic unless the physician explicitly forbids it (the ‘opt-out’ model).
• France: Allows for automatic substitution under specific conditions, primarily for hospital use.
• Italy: Encourages the use of biosimilars through regional tenders and prescribing targets but does not mandate automatic substitution at the pharmacy level.
• United Kingdom (post-Brexit): The MHRA has introduced legislation to allow pharmacists to automatically substitute biosimilars without needing to consult the prescriber, a move designed to increase uptake and cost savings.

For developers, this means that while the EMA marketing authorisation is the first hurdle, engaging with national payers and understanding local substitution policies is critical for market access. The lack of a centralised ‘interchangeable’ label simplifies the EMA assessment but complicates the commercial landscape.

Real-World Evidence (RWE) and Pharmacovigilance

Post-authorisation, the focus shifts to pharmacovigilance. The EU requires the inclusion of biosimilars in the risk management plans (RMPs) of the reference product. This ensures that safety data is pooled and monitored. Increasingly, Real-World Evidence (RWE) from registries is being used to monitor the impact of switching in large populations. For example, the DANBIO registry in Denmark and the BSRegistry in the UK have provided valuable data reassuring that switching from originator to biosimilar infliximab does not lead to increased immunogenicity or loss of efficacy.

Regulators prioritise RWE that is generated using robust methodologies (e.g., prospective data collection, appropriate confounder adjustment) to complement the clinical trial data. This is particularly relevant for biosimilars of complex molecules like insulin or long-acting granulocyte colony-stimulating factors (G-CSFs), where subtle differences in formulation might impact patient experience.

Regulatory Strategy: What Regulators Prioritise in Practice

From the perspective of an AI systems practitioner and regulatory analyst, the successful biosimilar development program is one that integrates data science and regulatory science from day one. Regulators prioritise specific aspects of the submission that signal high quality and low risk.

1. The Quality-by-Design (QbD) Approach

The EMA strongly encourages a QbD approach, where the product and process are understood and controlled based on sound science. A submission that presents a clear design space—the range of process parameters within which the product quality is assured—is viewed favourably. This demonstrates that the developer understands the critical quality attributes (CQAs) and how process parameters affect them. A ‘trial-and-error’ approach to process development is a red flag.

2. State-of-the-Art Analytical Techniques

Regulators are increasingly sophisticated in their understanding of analytical technology. Submissions that utilise advanced techniques such as high-resolution mass spectrometry, nuclear magnetic resonance (NMR), or advanced glycan analysis are expected. Simply using standard pharmacopoeial methods is often insufficient to demonstrate high similarity. The agency expects the developer to push the boundaries of analytical capability to reduce the residual risk.

3. Justification of Differences

No biosimilar is identical to its reference product. The EMA expects to see differences in the data. The key to success is not the absence of differences, but the robust justification that these differences are not clinically meaningful. This requires a deep scientific rationale, often supported by bridging studies or modelling. If a developer tries to hide or downplay analytical differences, the assessment will likely fail.

4. The Interchangeability Package (Voluntary)

Although not a regulatory requirement for EMA approval, some developers are voluntarily conducting additional studies to support interchangeability claims, particularly in anticipation of national substitution policies or to address prescriber concerns. These studies often involve multiple switches between the products (e.g., three-way switches) to detect any potential immunogenicity signal that might arise from alternating products. While the EMA does not mandate this, it is a strategic move to secure market share in competitive environments.

5. Data Transparency

The EMA places a high priority on transparency. The full data package, including clinical study reports, is often made publicly available. Developers must be prepared to defend their data and methodologies openly. The use of AI and machine learning in data analysis is growing, but regulators will scrutinise these models to ensure they are validated and do not introduce bias. For instance, using AI to analyse glycosylation patterns is acceptable if the algorithm is transparent and validated against reference standards.

Comparative European Landscape: Country-Specific Priorities

While the EMA provides the scientific assessment, the commercial reality is dictated by national payers and regulators. Understanding these nuances is essential for a comprehensive regulatory strategy.

Benelux and Nordic Countries: These regions have high uptake of biosimilars, driven by tendering systems and strong central guidance. In Sweden and Denmark, the use of registries to monitor outcomes is deeply embedded in the healthcare system. Regulators in these countries prioritise cost-effectiveness and real-world safety data. A developer focusing on these markets should prepare robust health economic models alongside the EMA submission.

Southern Europe (Spain, Italy): Uptake has historically been slower but is accelerating. The focus here is often on hospital formulary inclusion. Regulatory interactions at the regional level are as important as national policy. Demonstrating supply chain reliability and educational support for hospital pharmacists is often prioritised.

Eastern Europe: Markets in this region are price-sensitive and often rely on tenders. The regulatory scrutiny might focus more on the ability to supply the market consistently at a lower price point rather than the nuances of immunogenicity data, provided the EMA approval is in place. However, pharmacovigilance infrastructure varies, so developers must ensure they have robust local safety reporting mechanisms.

Post-Brexit UK (MHRA): The UK has diverged slightly by introducing the ‘abridged’ pathway for international recognition, allowing for faster approval of biosimilars already approved by stringent regulatory authorities (SRAs). However, the MHRA maintains the same scientific standards as the EMA for comparability. The UK’s push for automatic substitution is a unique feature that developers must factor into their launch strategy.

Emerging Complexities: Biologics Beyond Proteins

The EU framework is currently being tested by the next generation of biologics: antibody-drug conjugates (ADCs), fusion proteins, and complex generics (e.g., liposomal formulations). The EMA is issuing specific guidance for these product classes, recognising that the standard ‘biosimilar’ paradigm may need adaptation.

For ADCs, the comparability exercise is twofold: the antibody must be similar, and the linker-payload must be similar, and the conjugation process must be controlled. This introduces significant analytical complexity. Regulators are currently prioritising the development of analytical methods capable of characterising the drug-to-antibody ratio (DAR) and the distribution of DAR species.

Similarly, for biosimilar insulins, the transition from animal-sourced to recombinant DNA technology in the past required a distinct regulatory approach. Today, developers looking to create biosimilars of newer insulin analogues face the challenge of demonstrating similar PK/PD profiles while managing the immunogenicity risk, which is lower than for monoclonal antibodies but still present.

The Role of AI and Data Analytics in Regulatory Strategy

As an AI systems practitioner, it is evident that artificial intelligence is becoming a tool for both the developer and the regulator. Developers use AI to optimise cell line selection and culture conditions to ensure consistent glycosylation profiles. Regulators use AI-assisted review tools to handle the vast volume of data submitted in an MAA.

However, the EMA remains cautious about the ‘black box’ nature of some AI algorithms. If a developer uses machine learning to predict clinical outcomes based on analytical data, the agency will require full transparency on the training data, validation sets, and algorithm logic. The regulatory strategy must include a ‘Validation of AI Models’ section if such tools are used in the development or assessment process. This is an emerging area of regulatory science where the EU is currently defining best practices.

Practical Steps for Regulatory Submission

For professionals preparing a biosimilar for the EU market, the following strategic steps are recommended based on current regulatory priorities:

1. Early Scientific Advice: Engage with the EMA (or NCAs) early in the development process. The Scientific Advice procedure is invaluable for confirming the acceptability of the proposed comparability exercise, the choice of reference product, and the clinical trial design.
2. Reference Product Sourcing: Ensure continuous access to the EU-sourced reference product. The EMA requires that the reference product used in the comparability studies be sourced from the EU (or a market with similar regulatory standards) to ensure that the comparison is valid for the EU population.
3. Pharmacovigilance System Master File (PSMF): Prepare a robust PSM