Anything You Can Do I Can Do (Better?): Demonstrating Biosimilar Interchangeability

FDA recently finalized its guidance document intended to assist protein biosimilar manufacturers in demonstrating interchangeability, Considerations in Demonstrating Interchangeability With a Reference Product.  Interchangeability, as of now, is nothing more than a pipe dream; no biosimilar has been declared interchangeable and therefore fully substitutable for its reference product.  But FDA is hoping that this guidance will help address that.  This guidance finalizes FDA’s expectations and updates the draft version of the guidance to provide more clarity.  Of note, this guidance document is directed specifically toward proposed therapeutic protein products, which include those products scheduled to transition from drug products to biologic products on March 20, 2020 under the “deemed to be a license”  provision of the Biologics Price Competition and Innovation Act.

The draft version of this guidance was published in January 2017 and triggered over 50 separate (mostly substantive) comments from a variety of stakeholders arguing the merits of each recommendation made in the guidance.  Some, for instance, suggested that interchangeability should be established individually for all conditions of use while others contend that extrapolation across intended uses is sufficient; similarly, some argued that immunogenicity in product switches must be robustly assessed while others contend that immunogenicity is a “hypothetical concern”.  Comments also provided line-by-line analyses of the text, often requesting clarity, as well as concerns about the specific testing parameters set forth in the guidance.

In the almost 2.5 years since the publication of the draft, FDA has made some changes but not many significant ones.  Indeed, even though the 2019 version clocks in at 7 pages less than the draft, the final version is largely the same as the draft.  Though much of the language has been revised, the two versions differ mainly in detail and clarity.  For example, while the concepts remain, terminology like “fingerprint-like” demonstrations of biosimilarity and “residual uncertainty” has been scrapped.  More substantively, the final guidance omits discussion of the specific analyses FDA expects for product presentation, design, and administrative characteristics.  For that reason, the Appendices also differ between the draft and the final, with the draft’s appendix detailing considerations in Comparative Use Human Factors Studies replaced by an example of switching study design in the final.

The meat of the guidance document remains largely unchanged.  As in the draft, the final provides an overview of the important scientific considerations in demonstrating interchangeability with a reference product.  A product is interchangeable if it is biosimilar to the reference product and can be expected to produce the same clinical result as the reference produce in any given patient.  To demonstrate interchangeability, the risks involved in switching from the reference to the biosimilar may not be any greater than the risk of solely using the reference product when a biological product is intended to be administered more than once to an individual.  To meet this standard, the guidance discusses the:

• Data and information needed to support a demonstration of interchangeability;
• Considerations for the design and analysis of a switching study or studies to support a demonstration of interchangeability;
• Considerations regarding the comparator product in a switching study or studies; and
• Abbreviated considerations for developing presentations, container closure systems, and delivery device constituent parts for proposed interchangeable products.

FDA also provides a list of factors that may be required for demonstrating interchangeability but emphasizes that the data necessary may vary based on the nature of the proposed interchangeable product, as not all factors may be necessary.  The product’s degree of structural and functional complexity may influence the data required.  Clinical experience with the reference product may also affect the data and information needed to support a demonstration of interchangeability.  In such cases, FDA will require a scientific rationale for extrapolation of data and other information to support interchangeability.

FDA applies a presumption that switching studies are required for demonstrations of interchangeability.  Again, when a sponsor believes that switching studies are not necessary, the sponsor must provide a justification for not needing such data to demonstrate interchangeability.  Switching studies should be designed based on clinical practice and risk-based treatment scenarios, and FDA therefore has outlined a flexible approach regarding study design.  Primary endpoints should assess the impact of switching on PK and PD (if available), as well as descriptively assess immunogenicity and safety.  Assessments of efficacy endpoints can be used as supportive information, as may postmarking data of a product previously licensed as biosimilar (rather than interchangeable).  The guidance provides significant detail on suggested study design and analysis.  The final guidance also reiterates the need for “bridging” studies should the study use a comparator that is a non-U.S. version of the reference product.

Though the initial publication of this guidance was welcomed by industry, its impact remains unclear.  In the years since the publication of the draft, FDA still has not determined that a single drug is interchangeable to its reference product.  There is no telling when FDA will license its first interchangeable product, and it is clear that reference product sponsors will not make acceptance of biosimilars – interchangeable or not – an easy feat.  But if this now-final guidance has its intended effect, reference product sponsors will not be able to cast doubt on the sameness of biosimilars anymore.  Interchangeability should make the notion that “anything a biosimilar can do, the reference product can do better” a thing of the past, but this is all conjecture until a sponsor can actually demonstrate interchangeability.