The U.S. Biosimilar Pathway Nearly Three Years Later

Erika Lietzan Laura Sim

Note: The views expressed in this article are those of the authors and do not necessarily represent those of his/her employer, GxP Lifeline, its editor or MasterControl, Inc.

The statutory pathway for approval of biosimilars in the United States1, established in March 2010 with the enactment of the Biologics Price Competition and Innovation Act of 2009 (BPCIA), represented the culmination of 10 years of stakeholder discussion of scientific and policy issues relevant to these products. Nearly three years after its enactment, however, FDA has not approved any biosimilars, debate about implementation of the statute continues, and many questions remain.

FDA first formally invited stakeholder comment on the new law at a public hearing in November 2010 and in written comments to the related docket. The agency then took a critical step in February 2012 when it released three draft guidance documents addressing certain threshold scientific and regulatory issues.2 FDA accepted comments on the draft guidances through mid-April and invited comments at a public hearing in May (and in the related hearing docket). At every stage, stakeholders who participated in the pre-enactment discussion - including industry, patient groups, trade associations, medical associations, non-profits, pharmacies, and pharmacy benefits managers - explained their positions on the legal, scientific, and regulatory issues.

Many stakeholders asked FDA to clarify what "fingerprint-like analysis" means, including the tests and measurements involved. Some supported use of fingerprint-like analyses to address residual uncertainties about structural differences.

This article discusses a number of key scientific and regulatory issues concerning the biosimilar pathway that have become the subject of greater focus during the past year. For each issue, it provides a brief overview of the positions taken by FDA in the draft guidances and by stakeholders in the related dockets.

Use of Analytical Testing

The BPCIA provides that a biosimilar application must contain data from analytical studies demonstrating that the biosimilar is "highly similar" to its reference product notwithstanding minor differences in clinically inactive components (unless the Secretary of Health and Human Services waives this requirement).3 The draft guidances state that analytical studies provide the foundation for the demonstration of biosimilarity. They advise biosimilar applicants to develop orthogonal, quantitative methods to distinguish differences in product attributes and note that it may be useful to compare quality attributes using a "fingerprint-like analysis" that covers a large number of attributes and their combinations. FDA previously discussed assessing a molecule's structure using a fingerprint-like analysis in connection with its November 2011 approval of an abbreviated new drug application (ANDA) for the low molecular weight heparin product enoxaparin sodium.

Many stakeholders asked FDA to clarify what "fingerprint-like analysis" means, including the tests and measurements involved. Some supported use of fingerprint-like analyses to address residual uncertainties about structural differences. Others took the position that the aggregation of metrics provides no more information than assessing metrics individually, and they urged FDA to remove references to fingerprint-like analyses from the final guidances. Stakeholders also disagreed about whether analytical testing or clinical testing should serve as the foundation of a biosimilarity demonstration.

Intentional Differences

The draft guidances indicate that biosimilar applicants should compare the primary and higher order structures of the two products but stop short of stating that the primary structures must be identical to meet the "highly similar" statutory standard. Minor modifications - such as N- or C- terminal truncations that will not affect safety, purity, or potency - may be permitted. Many stakeholders took the position that the primary structures should be identical and that the higher order structures should be as similar as possible. Some seemed comfortable with differences in N- and C- terminal truncations, if justified. The draft guidances also hint at the possibility of differences "introduced by design" and give the example of PEGylation. Many stakeholders took the position that intentional differences between the products should be minimized to the extent possible. Concerns were voiced in particular that structural differences may introduce clinical differences that may go undetected.

The statute requires that the route of administration, dosage form, strength, and conditions of use of the biosimilar be approved, also, for the reference product. The draft guidances suggest that FDA will be open to certain differences in other product aspects, provided the product still meets the approval standard of biosimilarity.4 For instance, FDA will accept some differences in formulation, delivery device and container closure system (provided adequate performance data are provided), excipients (although additional toxicity studies may be necessary), and primary packaging. Many stakeholders took the position that FDA should exercise caution with respect to, or even prohibit, differences that are reasonably avoidable. A few opposed differences in formulation, and one suggested that differences in formulation should not be permitted if new risk information would be needed in the labeling. A few stakeholders suggested that delivery device and container closure differences should preclude a finding of interchangeability, and one raised the question whether FDA would permit promotional claims of superiority with respect to convenience or compliance where these features had been changed.

Bypassing Animal Testing

Under the BPCIA, a biosimilar application must include data derived from animal studies, including assessment of toxicity (unless waived by the Secretary).5 The draft guidances state that animal toxicity studies are useful when, after structural and functional characterization, uncertainties remain about safety that need to be addressed before the initiation of clinical studies. They therefore suggest that in some cases the statutory requirement for animal toxicity studies may be waived.

Commenters were roughly divided between whether animal testing is generally (or always) warranted or rarely (or never) warranted. Some stakeholders indicated that the need for animal testing depends on whether a relevant species exists. Others stated that nonclinical toxicology testing, even in a non-relevant species, is valuable before testing a new biological product in humans.

Clinical Study Design

A biosimilar application must include data derived from a clinical study or studies, including the assessment of immunogenicity and pharmacokinetics or pharmacodynamics, that are sufficient to demonstrate safety, purity, and potency in one or more appropriate conditions of use (unless waived by the Secretary).6 The draft guidances provide that at least one comparative clinical immunogenicity study is generally expected and that it is generally important to demonstrate only that immunogenicity is not greater with the proposed biosimilar than with the reference product (i.e., a one-sided study design is generally sufficient). The minimum follow-up period for chronically administered agents is one year, unless a shorter period is justified. The draft guidances also state that comparative safety and efficacy data are necessary if residual uncertainties about biosimilarity remain (after completion of earlier testing) and that in some cases a one-sided design may be appropriate to evaluate efficacy.

Many commenters asserted that the clinical evaluation of immunogenicity is essential to any biosimilarity demonstration, though some argued that this testing often will not be warranted. Some stakeholders supported one-sided study designs while others urged FDA to require two-sided study designs (because less immunogenicity could suggest that other clinically meaningful differences between the products may exist). A few stated that the one year of follow up should take place preapproval.

Stakeholder views on the design of clinical efficacy studies also varied, with some supporting a one-sided design for efficacy studies and others favoring a two-sided study design to detect whether a proposed biosimilar is more effective than its reference product - in which case, they argued, the product should be licensed under the pathway for full applications, section 351(a) of the Public Health Service Act (PHSA). Several stakeholders argued that the endpoints that supported licensure of the reference product should be studied, whereas others asserted that the endpoints most sensitive to potential product differences should be evaluated.

Role of Non-U.S. Comparator Data

The extent to which a biosimilar manufacturer may rely on comparative data involving a product not licensed in the United States (a non-U.S. comparator product) has been the subject of considerable discussion since FDA raised the issue in late 2010 and especially since the draft guidances were released. The draft guidances take the position that, as a scientific matter, analytical studies and at least one human pharmacokinetics (and if appropriate one pharmacodynamics) study must involve the U.S.-licensed reference product. They state, however, that a biosimilar applicant may rely on comparative animal or clinical data involving a non-U.S. comparator product, if adequate justification is provided.

Some stakeholders questioned the legal basis for the reliance on non-U.S. comparator product data, given that the BPCIA provides that "reference product" means "the single biological product licensed under [PHSA § 351(a)] against which a biological product is evaluated" in a biosimilar application.7 Some argued that there will rarely (or never) be scientific justification for this reliance, whereas others stated that this reliance is generally appropriate and will reduce duplicative testing. Many requested more information about the design of bridging studies needed to support use of non-U.S. comparator product data.


The statute requires a biosimilar product to be proposed for conditions of use that have been approved for the reference product.8 The draft guidances indicate that FDA will permit applicants to demonstrate biosimilarity with respect to one reference product indication and, with appropriate scientific justification, to be licensed for another reference product indication. The draft guidances suggests factors that must be considered in that justification, including whether the mechanism of action, pharmacokinetics and biodistribution, and expected toxicities are the same in the two indications and populations. FDA also recommends testing in the "most sensitive" population, noting that patient populations may have different comorbidities and receive different concomitant medications.

Many stakeholders supported extrapolation where scientifically justified and agreed with the agency's list of considerations and emphasis on a "most sensitive" population. Indeed, some took the position that extrapolation should be automatic. Others, however, were more cautious. A few took the position that extrapolation should not be permitted, or should not be permitted with respect to immunogenicity (i.e., immunogenic risk should be assessed in each relevant patient population). A few noted that as a practical matter, there may not be one single "most sensitive" population.


The scientific and regulatory issues discussed above need to be resolved for section 351(k) to be a workable pathway for applicants. Many other scientific and regulatory questions also need to be resolved, including whether biosimilars will bear distinctive non-proprietary names, how they will be labeled, on what basis they will be designated interchangeable, and what sort of post-market commitments will be imposed. Some very important legal questions remain, including how FDA will interpret the exclusivity provisions of the statute and whether it will ensure that biosimilar applicants provide their applications and manufacturing information to reference product sponsors.

Stakeholders would benefit from a transparent and public, but swift, resolution of these issues. FDA has indicated it will finalize the draft guidances and continue to issue new draft guidances. Regulations on some topics may eventually issue, although the agency has not indicated plans to take up rulemaking, and it was slow to issue regulations following the generic drug amendments in 1984. It is thus possible that the agency will begin approving biosimilar products before its views on many important legal, scientific, and regulatory issues have been shared with the broader industry and the public.


  1. A biosimilar is a biologic approved on the basis of a truncated application that refers to a previously approved ("reference") product and contains (presumptively) an analytical, preclinical, and clinical comparison with that reference product.
  2. See "Scientific Considerations in Demonstrating Biosimilarity to a Reference Product," 77 Fed. Reg. 8883 (Feb. 15, 2012); "Quality Considerations in Demonstrating Biosimilarity to a Reference Protein Product," 77 Fed. Reg. 8884 (Feb. 15, 2012); and "Q & As Regarding Implementation of the BPCI Act of 2009," 77 Fed. Reg. 8885 (Feb. 15, 2012). Although the substance of each guidance differs, the topics they address often overlap, and this article refers to them collectively for simplicity.
  3. Public Health Service Act (PHSA) § 351(k)(2)(A)(i)(I)(aa).
  4. PHSA § 351(i)(2) (defining "biosimilar" or "biosimilarity" as meaning that "the biological product is highly similar to the reference product notwithstanding minor differences in clinically inactive components" and "there are no clinically meaningful differences between the biological product and the reference product in terms of the safety, purity, and potency...").
  5. PHSA § 351(k)(2)(A)(i)(I)(bb).
  6. PHSA § 351(k)(2)(A)(i)(I)(cc).
  7. PHSA § 351(i)(4).
  8. PHSA § 351(k)(2)(A)(i)(III).

Erika Lietzan is a partner in the food and drug group at Covington & Burling LLP. She specializes in the regulation of drugs and biological products. Her areas of special focus at this time include biologics and biosimilars; the Hatch-Waxman amendments; data and market exclusivities and lifecycle strategy; regulation of clinical trials; supply chain issues; and post-market drug safety. She speaks and publishes regularly on these topics. Ms. Lietzan works with a number of trade associations, as well as many of the country's major biopharmaceutical companies and many smaller companies. Among other things, she assists in regulatory, legislative, and policy advocacy in the U.S. and around the world; provides long-term strategic advice with respect to regulatory policy and product lifecycle strategy; prepares clients for regulatory meetings and legislative hearings; counsels and advocates in the course of criminal and congressional investigations; and serves as an expert in litigation. For two years, Ms. Lietzan was assistant general counsel of PhRMA. She is past chair of the Biotechnology Committee of the American Bar Association, and she is currently a member of the FDLI Board of Directors and an elected member of the American Law Institute. Ms. Lietzan can be reached at or (202) 662-5165.

Laura Sim is a senior associate in Covington & Burling LLP's food and drug group. She advises drug, biologic, and food manufacturers and trade associations on a range of regulatory and consumer protection issues, with a focus on product life cycle management, biosimilar regulation, drug safety, and the intersection of the drug, medical device, and food industries with laws administered by the Consumer Product Safety Commission. Ms. Sim can be reached at or (202) 662-5262.