When more than $10 billion worth of branded biopharmaceuticals 1 go off patent in the next few years, as is currently projected, will any of them face generic competition in the U.S.? Despite the high cost of these products to patients,2 they will not, unless there is legislative and regulatory change. But change appears to be coming.
In the past year, FDA has solicited written comments and convened two public workshops to discuss scientific issues at stake in such approvals. In addition, last summer, Congress held an initial hearing on the subject of generic biologics, though it has ceded responsibility for leadership of the current debate to FDA. In this role, FDA has been actively collecting input from pharmaceutical companies, biotechnology companies and other interested stakeholders to determine whether it is possible to approve a biogeneric as safe and effective for human use in the absence of the full complement of pre-clinical and clinical trial data currently required for all biological products; and if such an abbreviated approval pathway is scientifically feasible, what type and extent of data should be required.
The Regulation of Drugs and Biologics
Unlike drugs,3 which are approved under the federal Food, Drug and Cosmetic Act (FDCA), biologics are, with few exceptions, approved under the Public Health Service Act (PHSA).4 Despite a recent reorganization in agency jurisdiction, which moved the approval and regulation of all therapeutic proteins to FDA’s Center for Drug Evaluation and Research (CDER), the approval of these products continues to be governed by the PHSA. This regulatory status has significant consequences for would-be biogenerics, because, unlike the FDCA, the PHSA offers no abbreviated route to approval.
Approval of entirely new products is comparable under the two acts. Under the FDCA, a new drug application (NDA) must include a full complement of pre-clinical and clinical data to establish the drug product’s safety and effectiveness. Similarly, under the PHSA, a biologics license application (BLA) must include complete reports of pre-clinical and clinical data establishing the biological product’s safety, potency 5 and purity.
Notably, however, for those products approved under the FDCA, there also exist two abbreviated routes to approval: the Abbreviated New FDCA Drug Application (ANDA) and the section 505(b)(2) application. Both of these abbreviated approval pathways were created by The Drug Price and Patent Term Restoration Act of 1984 (Hatch-Waxman Amendments). These abbreviated pathways do not, however, apply to biologics approved under the PHSA.
Arguably, for the small set of biologics such as human growth hormone and insulin, which for historical reasons are approved under the FDCA, the law currently provides sufficient authority to permit an abbreviated approval procedure for generic versions of these products.6 For the vast majority of biologics, however, creating such a pathway would require legislative action because currently FDA does not have jurisdiction to do so under the PHSA.
Significantly, moving toward legislative and regulatory change requires resolution of a set of difficult scientific and technical questions—questions on which FDA is currently gathering information. Indeed, at last summer’s hearing before the Senate Judiciary Committee, when Senators Orrin Hatch (R-UT) and Charles Schumer (D-NY) pressed FDA Acting Commissioner Lester Crawford on why FDA’s actions on this issue had been so delayed, the acting commissioner explained that FDA was working to resolve the scientific issues before taking up an appropriate regulatory scheme.7 Evidently, at least some members of the Senate may also be now leaning in this direction, for as Stephen Northrop, director of the Senate HELP Committee’s health policy staff, recently indicated, the committee itself wants to have a good understanding of the scientific issues and FDA’s findings with respect thereto so that it can determine how best to address the legislative issues.8
Models for Abbreviated BLA Approval
To begin crafting the broad contours of an abbreviated approval pathway for biopharmaceuticals, some have turned to the FDCA for possible models or, more specifically, to the ANDA and the 505(b)(2) abbreviated application. Whereas the ANDA allows FDA to approve a generic drug that is the “same as” a previously approved brand product, Section 505(b)(2) permits abbreviated approval of a drug that is somewhat less strictly identical, provided the scientific literature or other data demonstrate that product’s safety and effectiveness. Many have expressed concern that the ANDA process may be unfeasible because of the inherent difficulties involved with a biologic satisfying the “same as” requirement of the ANDA process codified in section 505(j) of the FDCA. Thus, the 505(b)(2) provides the more likely model of the two, though it is not without problems. A brief overview of each of these abbreviated approval pathways for drug products approved under the FDCA, and their potential applicability to biogenerics, is provided below.
Abbreviated New Drug Application (ANDA)
Of the two abbreviated approval routes available under the FDCA, the less likely model for biogenerics appears to be the ANDA, permitted only where the applicant shows that its drug product is the “same” as the approved drug on which it is modeled (known as the “reference listed drug”). Sameness is demonstrated by showing that the proposed drug:
| • | Has the same active ingredient, strength, dosage form, route of administration, labeling and conditions of use; and |
| • | Is bioequivalent to the reference listed drug. |
Having demonstrated that its product is the “same” as the reference listed drug, an applicant need not submit studies establishing its own drug’s safety and effectiveness.
The sole exception to the sameness requirement for purposes of ANDA eligibility is in the case of an FDA-approved “suitability petition.” A suitability petition is permitted only where the differences between the proposed and the listed drugs are of a kind requiring no additional clinical data: changes of dosage form, route of administration, strength, or, rarely, active ingredient.9 Any other differences between a proposed drug product and the reference-listed drug require the applicant to submit a new drug application (NDA), though possibly under the abbreviated mechanisms of Section 505(b)(2) of the FDCA (which is further explained below).
Finally, once sameness and bioequivalence have been established, the generic version is deemed to be therapeutically equivalent to the reference-listed drug. That determination reflects the expectation that the generic version will produce the same clinical effect and safety as the reference listed drug, and thus allows the two drug products to be treated as interchangeable.
According to critics, the ANDA cannot provide a model for follow-on biologicals because, as they argue, the active ingredients of biopharmaceuticals are so complex that they cannot be fully characterized. Thus, it is impossible to prove that the follow-on product has the “same” active ingredient(s). As between the ANDA and the 505(b)(2) processes, the 505(b)(2) appears to offer a more viable model.
505(b)(2) Application
The key difference between the ANDA and 505(b)(2) approval routes is the degree to which the newly proposed drug differs from the reference listed drug. Where the ANDA requires proof of “sameness” (or an FDA-approved suitability petition for very limited types of differences), the 505(b)(2) application may involve a product with more significant differences. A 505(b)(2) application must include all pre-clinical and clinical data necessary to establish the safety and effectiveness of the drug product, notwithstanding these differences. It is in this regard that a 505(b)(2) application is most like an NDA; that is, the application must contain full reports of investigations of safety and efficacy. Unlike full NDAs, though, 505(b)(2) applicants may take some of that safety and efficacy data from studies “neither conducted by the applicant nor for which the applicant has obtained a right of reference.”10 FDA interprets this statutory language to mean that a 505(b)(2) applicant may rely not only on published studies or data otherwise in the public domain, but also FDA’s own prior finding of safety and effectiveness for a previously approved drug, even where those findings are based on the proprietary data of the submitter of the NDA upon which the 505(b)(2) relies.11
What makes the 505(b)(2) a somewhat more likely model for follow-on biologicals is the relaxation of the “sameness” requirement so as to allow differences between the applied-for and reference listed drug products, so long as all necessary data are submitted to establish the safety and effectiveness of the product. In the case of a 505(b)(2) application for a generic biologic, for example, the areas of comparability, if any, sufficient to permit reliance on FDA’s prior approval of the reference-listed drug would first be identified. In those areas where sufficient comparability is lacking, the 505(b)(2) applicant would be required to submit its own supplemental pre-clinical and/or clinical data necessary to ensure the product’s purity, potency, and safety.
Yet the increased flexibility of the 505(b)(2) route does little to resolve the many scientific issues that have been raised concerning follow-on biologics. These scientific issues are explained below.
Follow-on Biologicals: Scientific Issues
For those closely following this evolving area of the law, perhaps the most eagerly anticipated event is the publication of FDA’s promised “scientific framework” guidance document.12 FDA has recently announced that this guidance may take the form of a series of guidance documents, each discussing one or more of the specific scientific issues raised in the debate concerning biogenerics—the first two of which FDA has indicated will most likely cover the topics of immunogenicity and chemistry. All of these guidance documents are anticipated to be issued in draft form and be made available for public comment before being finalized. FDA’s primary focus in these guidances will be on defining how it handles issues related to biological comparability across different regulatory structures and in different stages of product lifecycles. In doing so, FDA hopes to ensure that the agency applies a consistent scientific approach across products, regardless of whether the various products are approved under the PHSA or the FDCA.
Although the scientific arguments raised in the debate concerning follow-on biologics are too numerous and detailed to permit a full explication of them here, their main contours are summarized below. The arguments, which arise from differences between chemically synthesized drugs and biologic products, can be loosely classified as deriving from one of two main premises: (1) that unlike chemically synthesized drugs, biologically derived products raise special concerns due to their size, complexity and heterogeneity; and (2) that there is an inextricable link between a biologic’s manufacturing process and its clinical attributes. Whether these challenges are surmountable, and/or whether the differences are overstated, are key questions in this important scientific debate.
 |
Because biopharmaceuticals are drugs derived from living organisms that are typically large, complex, heterogeneous molecules, they are more difficult than chemically synthesized drugs to consistently manufacture, quantify and purify. The molecule of a therapeutic protein—typically composed of anywhere from tens of thousands to millions of atoms—is several hundred times the size and complexity of a chemically synthesized drug. Moreover, proteins are composed of linear arrays of up to 20 different amino acids and can range from several dozen to several thousand amino acids in length.13
In addition, the overall shape of the protein contributes to its bioactivity. Its shape, in turn, depends on its amino acid sequence. In some proteins, however, the same linear string of amino acids may be folded differently in different organisms.14 Certain seemingly subtle changes in a protein’s structure can alter its bioactivity. In addition, many biologicals undergo complex post-translational modifications. One such type of post-translation modification that has been receiving much attention in the debate over follow-on biologics is that of glycosolation. Glycosolation is the addition of certain sugar molecules to the protein that cause the protein to fold in a unique way. How the protein is folded determines its function and whether it is active or not. Glycosylation is a critical step for those proteins that require an effector function, such as monoclonal bodies used to trigger an immune response in humans.15
To clarify, the active ingredient of a chemically-synthesized drug can be readily characterized, with precision, by widely accepted physical and chemical assays. Thus, any differences between a reference listed drug and a generic drug are easily determined. Such determinations become more challenging—some say impossible—in biological products.
 |
Process and Product: Inextricably Linked?
Often framed by the phrase “the product is the process, and the process is the product,” it is argued that unlike the clear and linear manufacturing process of a chemically synthesized drug, biological products have an inherent metabolic and synthetic variability that may cause small differences between manufacturing processes to result in significant differences in the clinical properties of the products. Some argue that each manufacturing process results in a different product, including different mixtures of active and inactive molecules and the levels of process- and product-related impurities. Small differences between manufacturing processes may cause significant differences in the clinical properties of the products, which—under this view—makes it impossible for different manufacturers to produce an identical biological product. And, as noted previously, the ability to produce identical products is of central importance in assessing whether to offer abbreviated approval for biopharmaceuticals.
Maintaining the stability of purified proteins is just one of many challenges. Since function depends on the protein’s three-dimensional properties, the configuration must be maintained during preparation, purification and storage. Moreover, as some argue, small structural changes introduced during the manufacturing of biologicals can affect a range of other parameters, including efficacy and immunogenicity. In addition, careful attention must be given to avoid unwanted microbial or viral contamination. Although the risk of introducing viruses that can infect humans is lower for manufacturers using non-mammalian host cells, all cell cultures require close attention to ensure that reagents, stabilizers and equipment used in the preparation of the product are free of contaminants.
The type of cell production system used to produce a biological product may have significant consequences for FDA’s ultimate conclusions with respect to these scientific issues, including the arguments made with respect to the difficulty or not of ensuring comparability between a previously approved biologic and its generic version. Some biologics are natural proteins extracted from animal or human tissue or blood, while others are manufactured in cells of non-human origin. In these latter recombinant technologies, a gene that codes for a certain human protein is inserted into the host cell. These host cell systems are of varying complexity—from microbial systems using e. coli and yeast as host cells, to mammalian systems using certain types of animal cells (e.g., Chinese hamster ovary cells) to, most recently, plant-based systems such as corn and tobacco. The choice of the host cell depends, in large part, on the final desired structure of the protein product.
Bacterial and yeast are the production systems of choice for relatively small proteins that do not require glycosolation or any other post-translational modifications. For such cultures, production costs are low and scale-up is relatively easy. Examples of biologics made with bacteria as the host cell include growth hormone, insulin and some interferons.16 Yeast as the host cell may be preferable, however, because it is not only relatively inexpensive to grow and scale-up, but it also eliminates concerns over harmful endoxins, of which proteins grown in bacteria cultures must be purified. Biologics produced with yeast as the host cell include, for example, glucagons and sargramostim.17
Mammalian cell culture is more technically complex 18 and more expensive than microbial cell fermentation.19 Unlike bacterial and yeast cultures, however, a mammalian culture system is able to manufacture glycosylated proteins that are required for clinical effect in humans. It is, therefore, usually only used in the manufacture of therapeutic proteins that show extensive and essential post-translational modifications such as glycosylation. In other words, the use of mammalian cell culture would be appropriate for those therapeutic proteins where the carbohydrate content and pattern is essential either to the protein’s biological activity, stability, and/or serum half-life. Such therapeutic proteins include, for example, erythropoietin and some cytokines and monoclonal antibodies.
A plant-based protein production system is an additional recombinant technology that is also being further researched and developed. In this process, plants themselves, such as corn and tobacco, become the factories that manufacture the proteins. These proteins are then extracted, refined and used in pharmaceutical production. One advantage of a plant-based system is that, like mammalian cell systems, plant cells can mimic human cells in certain ways. In many cases, the proteins are folded and cross-linked correctly and are biologically active. Notably, however, plant cells glycosylate differently from mammalian cells, which may, depending on the protein manufactured and its intended use, raise immunogenicity and/or efficacy issues.20
Although it is estimated that it may take another three to five years before full commercialization of the first plant-made pharmaceutical is achieved, the plant-based protein production system appears to be viable in the years to come. In particular, there is general consensus that while it is unlikely that plant-based biopharmaceuticals will replace bacterial fermentation or mammalian cell culture technology, they may provide a cost-efficient way to produce those proteins, such as monoclonal antibodies and certain interferons, which will be required in very large quantities.21
Forecast for Biogenerics in the U.S.
Generic versions of those biologics currently subject to the approval provisions of the FDCA will likely be among the first to be approved, but it will be quite some time before an abbreviated approval pathway is put into place for most biologics. Given the scientific concerns, any legislation implementing an abbreviated approval pathway for those biologics approved under the PHSA will likely be drafted broadly so as to permit FDA to adjust the required data package on a case-by-case basis. Even some staunch supporters of generic competition have advocated a case-by-case approach as most appropriate because of the wide-ranging complexity of biopharmaceuticals.22 Such an approach will likely begin with relatively non-complex, non-glycosolated, highly purified molecules for which the mechanism of action is well understood and then incrementally work toward developing an abbreviated approach for more complex therapeutic proteins such as erythropoietin and certain other glycoproteins. In short, there is, for follow-on biologics, not likely to be legislation created that contemplates a one-size-fits-all approach.
Other questions to be addressed through legislation include whether to provide for two separate tiers of data requirements depending upon whether an applicant seeks only FDA approval for its biogeneric or whether the applicant also seeks a rating of therapeutic equivalence that will permit the generic version to be deemed interchangeable with the reference listed biologic. Still other questions include whether the abbreviated approval pathway for biogenerics should include a patent-certification process similar to that devised by the Hatch-Waxman Amend- ments. For example, Congress will need to determine whether to adopt the Hatch-Waxman approach of permitting would-be generics to challenge patents. If so, questions concerning the types of biopharmaceutical patents that should be eligible for Orange Book listing, including the question as to whether process patents should be eligible, will need to be addressed. Also, questions will need to be answered concerning whether to apply certain non-patent market exclusivities to the reference listed product so as to ensure appropriate incentive for the continued research and development of new biopharmaceuticals. If so, then the specific types and mechanics of such exclusivities will also need to be deliberated.
Finally, it is important to note that availability of bulk material and bulk suppliers must also be in place in order for a viable U.S. biogenerics market. Although several contract pharmaceutical producers have emerged during the past few years to service the biopharmaceuticals market generally, it is not at all clear what role, if any, these contract producers will play in the biogenerics market specifically.
In the short term, however, all eyes are turned toward FDA’s promised “background document” and scientific guidances. If the agency continues its pattern to date, more public hearings on the scientific questions can be anticipated. Moreover, it is likely that only after FDA has issued more definitive findings on the scientific questions that the arduous legislative process will begin.
