BIO Comments on FDA Biosimilars Guidance: Demand More Testing, Permit Fewer Differences

[author: James DeGiulio]

FDAIn February, the U.S. Food and Drug Administration published three draft guidance documents on biosimilars and the abbreviated biosimilar approval pathway created by the Biologics Price Competition and Innovation Act of 2009 ("ABLA pathway") (see "FDA Publishes Draft Guidelines for Biosimilar Product Development").  The FDA provided a 60 day period for stakeholders to provide comments on the guidances, opening a docket for the each of the three draft guidance documents, FDA-2011-D-0602 (Quality Considerations), FDA-2011-D-0605 (Scientific Considerations), and FDA-2011-D-0611 (Questions and Answers).  The FDA received a submissions from a wide array of stakeholders, including the Biotechnology Industry Organization (BIO), which published its comments on BIO's website on April 16.  BIO filed a separate comments document for each of the three guidance documents, first introducing the organization and thanking the FDA for the opportunity to comment on the guidances before launching into the group's positions.  Both the Scientific Consideration and Quality Considerations comments contain an index that proposes line-by-line changes to the FDA guidance language, which will not be discussed here.

Biotechnology Industry Organization (BIO)As an organization representing many innovators with currently available reference products (RP), BIO advocates that the FDA expand the requirements to prove biosimilarity significantly beyond the minimum tests required in the statute, and generally recommends policies that make it more difficult to obtain ABLA approval.  Business motives aside, BIO presents a persuasive and scientifically sound justification for each of its positions on how the ABLA pathway should be administered moving forwad.  Biosimilar applicants would be hard-pressed to argue that the minimum study requirements listed in the statute provide all the data needed to establish a safe and efficacious biosimilar product in most situations.  Overall, BIO commends the FDA for its first effort -- but makes it clear that the U.S. biotech industry still needs substantially more guidance before the ABLA pathway can be endorsed as clear and safe.  There can be no disagreement that the FDA is tasked with challenging policy decisions:  ensuring patient safety with sufficient biosimilarity testing, yet somehow keeping the costs of biosimilar approval low enough to encourage applicants to choose to file an ABLA over a BLA, which currently represents an infinitely more predictable approval route.

Scientific Considerations Guidance

The first and most lengthy of BIO's comments covers the Scientific Considerations guidance, and was authored by Kelly Lai, the Director of Science and Regulatory Affairs at BIO.  To open the comment, BIO notes its extensive experience with biosimilar products, referencing the group's benchmark document BIO Principles on Follow-On Biologics, which was sent to key legislators in 2007 and emphasizes the required balance between patient safety and incentives to innovate that is required of a biosimilar approval pathway.  Of note, BIO also requests that FDA publish future biologic product-specific guidances that include specific statistics and/or ranges acceptable for a determination of biosimilarity.  This is currently the model followed under the European Medicines Agency (EMA) biosimilar framework, which has published several product guidances, including most recently on monoclonal antibodies.

BIO recognizes the need for distinguishing biologics and biosimilars from traditional small molecule generic drugs, both in terms of complexity and the quantity of clinical data required for approval.  Of course, noting particularly that they are merely similar, not identical, like generic small molecule drugs.  The FDA has introduced a number of distinguishing factors proposed by BIO in previous agency submissions, but BIO lists additional key issues that must be evaluated during biosimilar approval.

The group voices several concerns with the FDA guidance as written.  Of particular concern to BIO is the use of the word "should" throughout the document.  Especially in the context of demonstrating biosimilarity using analytical or animal/human clinical studies, the FDA generally uses the word "should" to indicate that something is suggested or recommended, but not required.  BIO takes issue with the use of this word where certain studies are fundamental and/or necessary for a determination of biosimilarity, as it falls short of conveying that these studies are required -- BIO prefers "is expected to" or "will need to," such as the key phrase:  "It is expected that the expression construct for a proposed biosimilar product will encode the same primary amino acid sequence as its reference product."

BIO asserts that the language of the guidance is inconsistent with the statute, in that it implies that clinical trials are only a "residual requirement" that is triggered if there are gaps in the analytical, PK/PD, and safety results.  Instead, BIO advocates for a more extensive clinical trial requirement for biosimilarity, requesting that the FDA necessarily require several additional studies beyond the statutory minimum, including animal toxicity studies and human clinical trials evaluating safety and efficacy.  BIO further advocates that the minimum clinical trials required by the statute, human PK/PD and immunogenicity human trials, be enhanced.  Due to the multitude of factors affecting immunogenicity, the trial must be of sufficient duration to assess the effects of immunogenicity on PK, biodistribution, safety, and efficacy, and a significantly laborious pharmacovigilence for all biosimilars, that must extend into the post-marketing period for at least one year (more for certain products).

The organization is extremely wary of the FDA's allowance of extrapolation of data to other indications and patient populations by the biosimilar applicant.  BIO suggests a "cautious approach," where the FDA accepts this type of data only if:  (1) the mechanisms and sites of action for both indications are very well understood; (2) there are no significant differences between PK and bio-distribution in patient populations; and (3) the clinical study in the original indication is highly sensitive to differences that may emerge from the new indication, including immunogenicity.  Even if these criteria are met, BIO suggests that a confirmatory clinical study be conducted on PK/PD and immunogenicity for all indications.  Further, the FDA's requirement that only "most sensitive" population be tested is insufficient, for defining such a population will often be impossible, especially if indications affect different organ systems in the human body.  Further, at times, one population may display the most immunogenicity, while another population shows the most efficacy.  Therefore, a single population study fails to address these two important safety issues.

A related issue is the FDA's explicit allowance of biosimilar approval for subsets of route of administration, presentations, and conditions of use.  BIO's main point here is prominent and efficient labeling of the biosimilar product, with an emphasis on the differences between the biosimilar and the RP.  Of course, an emphasis on the differences in the product will have the effect of discouraging patient switching.  However, the point is well taken with regard to a lack of identity of these various product properties, and the FDA may avoid a finding of interchangeability of product property identity is lacking.  As a good example of the multitude of scenarios the FDA has yet to identify, BIO asks for clarification as to what level of supplementation will be required to add new indications to the ABLA applications.  In particular, the scenario where the BLA holder secures approval for a new indication for the RP, and the biosimilar applicant wishes to add that new indication to its label.

Quality Considerations Guidance

The second comments document on Quality Considerations guidance was authored by Andrew J. Emmett, Managing Director of Science and Regulatory Affairs at BIO.  BIO points out that all biosimilars must be evaluated in the context of all that is in the public domain regarding aspects of the manufacturing process.  However, BIO asks for additional clarification regarding the FDA's scientific and technical expectations, including clarity regarding the term "fingerprint-like analysis."  Indeed, this term is repeated throughout the guidance, referencing mysterious "additional product attributes" and their correlation to clinical safety and efficacy, consistency in manufacture, or some other meaningful feature for establishing biosimilarity.  The "fingerprint-like analysis" approach is discussed by the FDA in (slightly) greater detail in Kozlowski et al. in the New England Journal of Medicine, but the term remains vague at best (see "FDA Looks to Multiple Sources, Including EMA Guidelines, in Developing Biosimilar Approval Standards").  The origin of this approach appears to be the EMA monoclonal antibody guidelines, although how it translates to overall biosimilarity evaluation remains to be seen.

An interesting (albeit not unexpected) position from BIO is that any intentional differences in host cell type, primary structure, formulation, or immediate package and those of the RP should disqualify the applicant for ABLA filing.  BIO advocates that any applicants introducing avoidable differences between the two products must seek approval using a BLA, otherwise the biosimilar applicant is unethically introducing the risk of undetected and clinically significant adverse events.  If differences between the biosimilar and the RP can be avoided, this is the best approach to control the proposed biosimilar's safety, effectiveness, and immunogenicity profiles.  With these unavoidable differences, BIO recommends placing the burden on the ABLA filer to affirmatively show that any design differences do not to result in a clinically meaningful difference between the biosimilar and reference product, which should require at least one clinical study.  This seems reasonable; biosimilar applicants certainly will expect that more differences between their product and the RP will lead to additional testing and cost.  Indeed, changes in formulation excipients, equipment, raw materials used in the manufacturing process for the active ingredient, the container closure system and the cold chain distribution system could potentially have an impact on safety and/or efficacy of the biologic.  BIO cited patient safety for this rigorous standard, which is valid, but an unmentioned benefit is the closer to identical the biosimilar product is to the RP, the likelihood improves that the biosimilar product is covered by one or more BLA holder patents.

Questions and Answers Regarding Implementation

The third comments document on the Q&A Guidance was authored by Sandra J.P. Dennis, BIO Deputy General Counsel, and is the shareholder's chance to respond to the policy changes implemented by the FDA following the first round of public hearings.  The comments on the Q&A Guidance is organized by individual questions.  While many of the issues have been discussed under the Scientific and/or Quality Considerations, there are several key points raised in the third comments set.

BIO disputes the FDA's definition of "protein," which was added to the PHSA in order to consolidate approval of all "proteins" via BLAs; inclusion of this term also forecloses the filing of New Drug Applications (NDAs) for polypeptide products that are longer than 40 amino acids.  BIO criticizes this 40 amino acid cutoff as arbitrary, and instead advocates a more flexible standard, including more consideration of the presence of higher order structure.  BIO advances the same critique of the 100 amino acid cutoff for defining a "chemically synthesized polypeptide," which is excepted from BPCIA coverage.

A major topic of interest at the 2011 BIO International Convention seminar on biosimilars was the access of confidential information by the ABLA applicant and the BLA holders (see "Docs at BIO: Session on Patent Litigation Tactics and Strategies for Biosimilars").  In its comments, BIO requested explicit recognition from the FDA that certain information contained within the BLA, including manufacturing facilities, must be treated as confidential.  BIO generally disfavors the comparison of a prospective biosimilar to a non-U.S. licensed product, but in particular, it raises significant concerns over the confidentiality of RP manufacturing sites, in that this (non-public) information will be critical in determining whether it is appropriate to rely on studies involving a foreign comparator product.  BIO points out that several key points must be remedied with regard to confidentiality/trade secret protection during all ABLA reviews.  A non-exhaustive list includes:  (1) FDA's regulations on disclosure of BLA information must be updated for consistency with the BPCIA, for several provisions of a BLA become public immediately; (2) confidential commercial and trade secret information in a BLA must be protected from intentional or inadvertent use or disclosure; (3) an ABLA application must stand on its own, in that FDA approval of ABLA must be based upon ABLA data and "publicly-available information" (redacted BLA "action packages" published by FDA) regarding the RP; and (4) procedures must be adopted to assure proper review of ABLA and avoid potential disclosure or reliance upon the reference BLA.  BIO adamantly notes that the BPCIA does not authorize or permit reliance on non-public information regarding the reference BLA, and the FDA must ensure the ABLA is absent any reliance on non-public BLA information (such as from IND or biosimilar development meetings).

Finally, BIO suggests that the FDA should require biosimilar applicants to certify compliance with section 351(l)(2) of the PHSA, which requires the biosimilar applicant to timely provide the reference product sponsor with the ABLA and such other information that describes the process or processes used to manufacture the biosimilar product.  Should a biosimilar applicant seek to opt out of the certification notice provisions by failing to provide the required information, the FDA should adopt a policy that it will refuse to file the biosimilar application.

As discussed, BIO's comments have the effect of rendering ABLA approval more difficult, both for the safety of U.S. patients, and for the competitive advantage of its BLA holder innovator members.  Regardless of these business interests, and enhanced level of proof of safety and/or efficacy for ABLAs adds credibility to the FDA's determination of biosimilarity for these complex and potentially under-characterized therapeutics.  To support these comments, BIO intends to provide testimony at the FDA public hearing on the draft guidances scheduled for May 11, 2012.

For both sides of the story, Patent Docs will present an analysis of the comments and positions suggested by biosimilar developers in a future post.