FDA Releases Draft Guidance on Comparative Analytical Assessment for Biosimilars

The FDA today released draft guidance on the design and evaluation of comparative analytical studies that are intended to support a demonstration of biosimilarity. A previous draft guidance document on statistical approaches to evaluate analytical similarity was withdrawn in June 2018.

The new guidance recognizes that analytical procedures have improved, and manufacturing science and production methods have advanced, such that there is an enhanced likelihood that a proposed product can be deemed highly similar to its reference through better targeting of the reference’s physiochemical and functional properties. For this reason, the FDA encourages the use of available state-of-the-art technology to prepare analytical studies that comprise part of a biosimilar’s application package. The FDA also encourages submission of comprehensive analytical data at the investigational new drug application stage.

The rationale for and the approach to comparative analytical assessments must be clearly described by the product’s sponsor, as this information can inform how much data will be required from animal and clinical studies. A meaningful assessment depends on the capabilities of available assays to assess features like a protein’s molecular weight, complexity, degree of heterogeneity, functional properties, impurities, and degradation. Identification of product-related variants, like storage conditions that lead to aggregation, should be included in analytical characterization.

The 3-dimensional conformation of the protein, which can be difficult to define precisely, is key; any differences should be evaluated in terms of potential effect on function and stability.

Evaluation of multiple lots is encouraged for the estimation of product variability, and the agency encourages sponsors to consult with the FDA to ensure that they select an appropriate number of lots.

In assessing differences between products, emphasis should be placed on orthogonal quantitative methods to definitively identify any differences in product attributes, and it may be useful, says the agency, to compare differences in quality attributes using a meaningful analysis algorithm.

Factors to consider when performing the comparative analytical assessment include the following:

• The expression system for the product, which is expected to encode the same primary amino acid sequence as the reference
• The manufacturing process
• Physiochemical properties, including the primary, secondary, tertiary, and quaternary structure; and posttranslational modifications
• Functional activities, including the range of relevant activities for the product
• Target bindings
• Impurities
• The reference product and applicable reference standards, as well as a comparison with non—US licensed comparators and consideration of publicly available information
• The finished drug product, including its formulation and packaging
• Stability under multiple stress conditions

The guidance highlights the fact that a full characterization of the reference, and consideration of publicly available information, will form the basis of understanding the reference, and lot-to-lot variability will be inherent to the reference. Sponsors should acquire multiple product lots of the reference and should include 6 to 10 lots in their assessments. Ideally, these lots will be derived from different drug substance batches to capture variability. Sponsors should be prepared to account for all lots acquired and characterized.

When sponsors use non—US-licensed comparators, they should provide comparative analytical data and analyses for all pairwise comparisons. Combining data from US and ex-US references would not be acceptable.

Sponsors should develop comparative analytical assessment plans and discuss their approaches with the FDA as soon as possible in the development process. The agency recommends that sponsors develop a risk assessment tool to rank the reference’s quality attributes in terms of potential impact on mechanism of action and function.

Appropriate analyses of comparative analytical data are also necessary, and factors that should be considered when determining how to evaluate data include:

• The nature of the attribute
• Distribution of the attribute
• Abundance of the attribute
• Sensitivity of the assay
• Types of attributes and assays
• Publicly available information

The FDA will evaluate the totality of the analytical data, and results of a given assay that do not meet prespecified criterial are not, on their own, preclusive of a demonstration of high similarity, the guidance points out.