On November 2 and 3, 2010, the FDA held its fi rst public hearing to obtain input on specific issues and challenges associated with the implementation of the Biologics Price Competition and Innovation Act of 2009 (BPCI Act), which established an abbreviated approval pathway for generic versions of biological products. According to the FDA, the purpose of the part 15 hearing was “to receive information and comments from a broad group of stakeholders, such as healthcare professionals, healthcare institutions, manufacturers of biomedical products, interested industry and professional associations, patient and patient associations, third party payers, current and prospective biological license application (BLA) and new drug application (NDA) holders, and the public, regarding implementation of the BPCI Act.” To that end, over 40 speakers representing a diverse array of biopharmaceutical companies, generic drug companies, trade groups, universities and patient advocacy groups presented prepared statements to a panel of FDA experts over the two days of hearings.
Although the FDA had solicited public comment on nine topics in particular, the majority of the comments from both the public and the FDA panel centered on fi ve key areas: 1) the type and amount of preclinical and clinical data necessary to establish biosimilarity, 2) the showing necessary to meet the heightened standard of interchangeability, 3) the naming conventions to be used for biosimilar products, 4) the extrapolation of data obtained in a clinical trial for one indication to support a biosimilar application for other approved indications of the reference product, and 5) the use of data from clinical trials conducted abroad to support a biosimilar application in the U.S. We have summarized below the key points for each of these issues that were raised at the hearings.
- Clinical Trials for Biosimilars
Over the course of the two-day hearing, the FDA panelists repeatedly asked presenters their views on the number, size and type of clinical trials that should be required to demonstrate biosimilarity to a reference product. In particular, the panelists solicited input on what the standard should be to meet the statutory requirement of having “no clinically meaningful differences.” Various comments by the FDA panelists indicated that the agency is grappling with striking the appropriate balance in setting a sufficiently high bar to ensure patient safety, but not so high that the biosimilar pathway will be rendered unusable. A spokesman for Novartis, commenting on the company’s own experience with biosimilars, stated that it currently takes, on average, about seven years and approximately $200 million to develop a biosimilar product.
There was general consensus among the speakers that some amount of clinical trials would be necessary to support a biosimilar application, at least for the time being. Speakers differed, however, on the amount and type of clinical trials the FDA should require. Not surprisingly, representatives of large pharmaceutical companies generally advocated larger and more sophisticated clinical trials while generic drug companies mostly favored fewer, more limited clinical trials. For example, spokesmen for Roche and Genentech suggested that rigorous head-to-head trials showing equivalence in safety and effi cacy between the reference product and the biosimilar product are needed. Further, the Roche-Genentech speakers pointed out the complex and multifaceted mode of action of therapeutic antibodies in particular, and asserted that safety and efficacy for that class of biologics can only be determined in rigorous clinical trials. At least one patient advocacy group, the Immune Deficiency Foundation, argued that antibodies should be excluded from the biosimilar pathway altogether, at least for the time being. A speaker for Merck Bioventures, who said that the company expects to have fi ve biosimilar programs in advanced development by 2012, likewise asserted that clinical trials demonstrating bioequivalence are needed as well as a noninferiority study, while the representative from PhRMA advocated comparative studies culminating in safety and effi cacy clinical trials. The speaker for Johnson & Johnson noted that immunogenicity cannot be accurately predicted without clinical trials. Also, Novartis asserted that the clinical trials would need to show no differences between the reference product and biosimilar in safety, purity and potency and would need to include a switching study, while Shire likewise argued that a multiple crossover clinical trial would be needed and that any differences in the biosimilar product’s safety or efficacy should be explicitly noted in the product label. On the point of switching studies, one FDA panelist noted that such studies would appear to be required by the statute itself. The representative of the Biotechnology Industry Organization submitted that clinical trials would be needed in all indications for which the reference product is approved. Amgen took the position that the fi rst FDA approval of a biosimilar product should always be for biosimilar status only, and that interchangeability status should not be granted without post-marketing data.
In contrast, the representatives for Teva Pharmaceuticals and for Watson Pharmaceuticals both argued that a single clinical trial would be adequate for all approved indications. Momenta Pharmaceuticals asserted that the clinical trial(s) should only need to demonstrate comparability, and that requiring biosimilar applicants to submit proof of statistical noninferiority or equivalence could be impractical. The Momenta spokeman also argued that switching studies should not be necessary.
- The Interchangeability Standard
During the hearing, FDA panelists asked numerous presenters to comment on when a fi nding of interchangeability would be appropriate. A number of patient advocacy groups, including the Global Healthy Living Foundation and the Coalition of State Rheumatology Organizations, asserted that it is not possible at this time to demonstrate that any biosimilar product meets the heightened standard of interchangeability. This position was echoed by the Pfi zer and BIO representatives. The Momenta representative, however, argued that technological advances will eventually make it possible to demonstrate interchangeability without any clinical data, although he did not specify exactly how that would be done. A speaker for Dr. Reddy’s suggested that it is currently possible to design a biosimilar that is “therapeutically indistinguishable” from the reference product. The representative of another patient advocacy group, the Immune Defi ciency Foundation, argued that a showing of interchangeability requires a showing that the biosimilar product produces the “exact same” clinical result as the reference product. J&J’s representative asserted that any showing of interchangeability must include data from switching studies, while Novartis argued that the FDA should require the standard for interchangeability to require at least the same amount of data that the FDA requires for product changes to the innovator product. Spokesmen for the Arthritis Foundation and the American Academy of Pediatrics both argued that automatic substitution of the interchangeable product for the reference product at the pharmacy level should not be permitted, while a speaker for CVS Caremark noted that the industry will need a mechanism for understanding any designation for a biosimilar that is not akin to the AB rating utilized under Hatch-Waxman.
The nonproprietary name that should be given to biosimilar and interchangeable biological products was vigorously debated at the public hearing. Patient advocacy groups including the Arthritis Foundation and the Global Healthy Living Foundation asserted that it was critical for each approved biosimilar product to have a unique nonproprietary name that would distinguish it from the reference product. Biopharmaceutical companies such as Pfi zer, Merck, Roche-Genentech and Amgen likewise all advocated for biosimilar products having unique nonproprietary names as did the Pharmaceutical Research Manufacturers Association (PhRMA).
In contrast, spokesmen for the European Generic Medicines Association, the Generic Pharmaceutical Association and Teva all argued that biosimilar products should have the same nonproprietary name as the reference product. A representative for CVS Caremark argued that a unique nonproprietary name was needed only in “exceptional circumstances,” while Medco Health Solutions’ spokesman suggested a distinguishing prefi x or suffi x would be appropriate for biosimilar products but not for those meeting the interchangeability standard. A representative of Express Scripts likewise argued that a unique proprietary name for biosimilars was not needed, and that adverse events could be tracked without it through existing pharmacy systems utilizing the National Drug Code (NDC) identifi er. One FDA panelist, however, pointed to a recent study that concluded the NDC alone is not adequate for this purpose and some unique naming is needed. Another panelist expressed an overriding concern with making sure that an approved biosimilar is not substituted at the pharmacy level as interchangeable where multiple follow-on biosimilars are approved without unique identifiers. The representative for the Coalition of State Rheumatology Organizations argued that, as a practical matter, patients do not remember lot numbers and there is a real risk that drugs get switched as patients switch health plans, so a unique nonproprietary name is the only way to address these issues.
The appropriateness of extrapolating data from a clinical study for one indication to support a biosimilar application for other approved indications of the reference product was of particular interest to the FDA and speakers alike. Numerous spokesmen, including those for the Global Healthy Living Foundation, Pfi zer, Roche-Genentech, and Shire, asserted that it would not be appropriate to extrapolate across indications. Meanwhile, BIO took the position that extrapolation would be permissible only where the mechanism of action was thoroughly understood. One FDA panelist asked the Roche-Genentech representative if submission of clinical data on the “two most diverse” approved indications would be sufficient to allow extrapolation across the remaining approved indications. In response, the Roche-Genentech spokesman cited the approval of Rituxan for treating both rheumatoid arthritis and cancer as an example why even that approach would not be feasible.
A representative for Watson, on the other hand, argued that data from a single, randomized, controlled trial should be extrapolated to all indications. Likewise, spokesmen for TEVA and Merck both asserted that no other studies should be required if it is clear the other indications involve the same mechanism of action. While J&J noted that extrapolation should be possible “in theory,” it argued that merely showing a common mechanism of action should not be sufficient.
- Foreign Studies
A significant amount of discussion during the hearing concerned the appropriateness of using data generated from clinical studies conducted abroad to support a U.S. biosimilar application. While few speakers advocated a bright line rule on the issue, there were differing viewpoints on when such data can be used. The Pfizer representative, for example, advocated that such data could only be used to support a U.S. biosimilar application where the “drug substance” (as opposed to drug product) in the U.S. and the foreign jurisdiction where identical. Otherwise, separate studies conducted in the U.S. should be required. The Merck representative offered a different view, taking the position that a biosimilar sponsor should have the burden of demonstrating to the FDA the relevance of data generated against a foreign approved reference product, which at a minimum would require a showing that the foreign study used the same patient population, the same dose, the same route of administration and was sufficiently robust.
The Teva representative, however, asserted that foreign-generated data should be accepted to support a U.S. application as long as the mechanism of action is the same. A speaker for Watson asserted that the FDA should accept non-U.S. trials conducted with a non- U.S. sourced innovator product to the extent the trials were conducted in ICH regions. Dr. Reddy’s spokesman advocated the FDA accepting studies conducted in another jurisdiction where there is a clear demonstration by the sponsor that the foreign reference product is similar to the FDA approved reference product and the FDA confirms that the quality of the clinical studies is acceptable. Novartis also advocated using foreign data to support a U.S. biosimilar approval. The EGMA argued that clinical studies should not be required in each jurisdiction if the reference products in the jurisdictions are the same. Hospira asserted that studies conducted in the EU should be accepted in the U.S. if bridging data regarding comparability and Pk/Pd studies supported it. At least one FDA panelist, however, noted that the statute itself referred only to comparison to a product approved in the U.S.
- Other Issues
Two other issues raised repeatedly by the FDA panel over the two-day hearing were pharmacovigilance and product “drift.” Specifi cally, the panel asked various speakers what degree of pharmacovigilance should be required for biosimilars. One patient advocacy group, the Arthritis Foundation, asserted that “the highest level of pharmacovigilance” is necessary for biosimilars. Likewise, Merck argued that pharmacovigilance is needed, but did not discuss the amount that should be required.
However, spokesmen for CVS Caremark and Medco argued that there should be consistency in the REMS requirements for innovators and biosimilars in order to make sure there are no added burdens on pharmacies associated with biosimilars. Teva took the position that postmarketing studies should only be required on an “as needed” basis.
With respect to product “drift,” the FDA panel was clearly concerned with how to maintain biosimilarity, and particularly interchangeability, between an approved biosimilar product and the reference product while taking into account the potential over time for “product drift.” The Teva representative advocated that the same rule applicable to innovators on this point should also apply to makers of biosimilars on this issue. And, when posed this question by the FDA panel, Hospira’s representative advocated that the biosimilar sponsor must stay within predetermined release criteria and the originator should likewise be prohibited from allowing product drift.
The recent public hearing conducted by the FDA concerning the implementation of the BPCI Act offered the first glimpse into the specific issues the FDA must resolve in order to implement the Act and the likely positions of various public and private entities on those issues. This is expected to be the first of several public hearings on the topic. Because the hearing was essentially an information gathering exercise for the FDA, the agency did not make any clear pronouncements of how it intends to implement the statute. Nevertheless, based on the questions posed by the FDA panelists and the comments submitted by key stakeholders, it is clear that the above-mentioned issues will be vigorously debated within the FDA and by the public as the implementation process moves forward.
Transcripts of the hearing will be available online at http://www.regulations.gov sometime in December. The FDA has indicated that it will accept further public comment on these issues through December 31, 2010, via electronic or written submissions. Electronic comments can be submitted to http://www.regulations.gov, while written comments can be submitted to:
Division of Dockets Management (HFA-305) Food and Drug Administration 5630 Fishers Lane, Room 1061 Rockville, MD 20852. Comments should include the following docket no. identifier: FDA-2010-N-0477