The FDA issued a guidance document “Development of Therapeutic Protein Biosimilars: Comparative Analytical Assessment and Other Quality-Related Considerations” (September 2025) (here) that focusses on the analytical approaches to be employed for comparative testing to demonstrate similarity between the proposed recombinant protein product and the reference (recognizing that the tools within the guidance could be applied to other protein products).
The analytical comparative assessment is a key component of demonstrating biosimilarity for a biologic:
“Analytical studies that demonstrate that the biological product is highly similar to the reference product notwithstanding minor differences in clinically inactive components.”
The ability to compare protein products is based upon the capability of the analytical techniques to characterize the physiochemical and biological properties where there is an understanding of the relationship between analytical technique, the protein functionality that is being characterized, and the therapeutic effect/mechanism of action associated with that functionality. For the attributes that are to be characterized, there needs to be an understanding of their criticality. As stated in the guidance:
“FDA recommends that sponsors develop a risk assessment tool to evaluate and rank the reference product quality attributes in terms of potential impact on the mechanism or mechanisms of action and function of the product.”
The risk assessment tool will consider attributes that impact clinical performance and any uncertainty associated with a quality attribute. The prioritization of attributes should be considered when establishing comparative quantitative data analysis criteria.
Upon prioritization of the attributes, comparative attribute data should be generated on those quality attributes; for example, on impurities, level of heterogeneity (e.g., from posttranslational modification), three-dimensional structure, purity, and activity on an appropriate number of lots of both the proposed product and reference. With such comparison studies, it is important to understand the lot-to-lot variability for the reference product versus the proposed product where the guidance states:
“Considering the inherent heterogeneity present in protein products and the expected lot-to-lot variability stemming from manufacturing processes, the Agency recommends that a sponsor include at least 10 reference product lots (acquired over a time frame that spans expiration dates of several years) in the analytical assessment to ensure that the variability of the reference product is captured adequately.”
Similar language is stated for the proposed product where there is a recommendation that the sponsor includes at least six to ten lots in the comparative assessment and that these lots represent the intended commercial manufacturing process.
The guidance includes a data analysis section where an approach is provided to demonstrate that the proposed product is highly similar to the reference product. This approach is the use of quality ranges (QR) for assessing quantitative quality attributes of high and moderate risk.
“Comparative analysis of a quality attribute would generally support a finding that the proposed product is highly similar to the reference product when a sufficient percentage of biosimilar lot values (e.g., 90%) fall within the QR defined for that attribute.”
“The QR should be defined as (𝜇𝑅−𝑋𝜎𝑅, 𝜇𝑅+𝑋𝜎𝑅), where 𝜇𝑅 is the sample mean and 𝜎𝑅 is the sample standard deviation based on the reference product lots.”
X is a multiplier to define the quality range; generally, three standard deviations are justified, but the applied value should consider the risk/criticality of the attribute that is being compared.
When differences are noted between the proposed and reference product, the guidance stresses the importance of orthogonal analytical techniques to further characterize and determine the nature of the differences for that attribute. The conclusion for the comparative assessment should consider the totality of data for each attribute as a failure to meet a comparative acceptance criterion for a particular attribute does not automatically preclude a demonstration that the proposed product is highly similar to the reference product
If you have any questions related to this guidance, or need guidance regarding its application in your firm, please reach out to Lachman Consultants at LCS@lachmanconsultants.com.
