Major biopharma companies and industry associations had their say.
The final comment period for the US Food and Drug Administration’s (FDA’s) draft guidance about how to conduct analytical evaluation of similarity in biosimilars (“Statistical Approaches to Evaluate Analytical Similarity”) closed on Nov. 21, 2017, and the comments received from various stakeholders by the agency were published on the website www.regulations.gov in early December.
The draft guidance, which was introduced in September 2017, outlines the information that biosimilar developers need to obtain about the structural/physicochemical and functional attributes of the reference product and how it can be used in the development of an analytical similarity assessment plan for a proposed biosimilar. The guidance document also provides recommendations for statistical approaches for evaluating analytical similarity.
Comments were received from a number of drug manufacturers, including Amgen, Boehringer Ingelheim (BI), Genentech, a Roche company, Novartis, Pfizer, Sanofi, and Shire, and industry associations, such as the Biosimilars Forum, the Biotechnology Innovation Organization (BIO) and the Association for Accessible Medicines (AAM).
Pfizer suggested that statistical analyses should not be a “default expectation” but used as a supportive tool, “when data are amenable to statistics and the statistical analyses results are meaningful to the understanding or interpretation of data”. Pfizer also noted that the FDA did not provide information on how to determine an appropriate biologically or clinically meaningful “margin for equivalence testing.”
While BI supports the idea of requiring an analytical similarity assessment plan, the company would like to see FDA apply scientific and regulatory consistency to all biologics and acknowledge the importance of prior knowledge of biologics (i.e. understanding quality attributes the molecular basis of diseases and the dynamics of vulnerable populations) for evaluating similarity.
Genentech commented that the FDA should include a requirement for protocols not just for statistical methods, but also experimental design. The company would also like to see information about the interactions FDA is willing to have with sponsors on protocol development and the milestone that needs to be met as part of the performance assessment.
AAM and the Biosimilars Council both support the draft guidance. However, AAM commented that one of the challenges that should be included for biosimilar developers is their lack of control of the quality attributes of reference products, which have changed over time for many different products on the market and can impact equivalence testing. The Biosimilars Council is concerned that strict pass/fail criteria for equivalence testing for Tier 1 attributes could, “have a substantial detrimental impact for development of biosimilars.”
The Biosimilars Forum recommended that FDA be willing to meet with biosimilar developers early on in the process to reduce misunderstanding. It also suggested the agency keep physical/chemical attributes and biological functions as separate concepts because different sponsors take different approaches with respect to what are considered to be product quality attributes.
