The January 2012 CMC (Chemistry, Manufacturing, and Controls) Strategy Forum focused on “Biosimilar Products: Scientific Principles, Challenges, Opportunities.” Traditionally, these forums describe the science and experience of various stakeholders in the pharmaceutical regulations process. The reports are reviewed and distilled to a consensus document. This distillate is designed to present the science related to a particular topic (problem). Ideally, this aids regulators and developers in adopting effective and economical science-based practices. So far, the track record has been supportive. The CMC Strategy Forum, held at San Francisco’s Intercontinental Hotel on January 22, 2012, focused on biosimilars (aka follow-on biologics). Despite broad public support, the race to follow-on biologics is going slowly due to very high, perhaps insurmountable, hurdles. These include legacy methods, proprietary specifications, lack of regulations (especially harmonized regulations), a byzantine legal process for sharing intellectual property in the U.S., lack of appropriate analytical standards and compendial methods, and how to handle naming of the product.
Dr. Cornelia Ulm opened the forum with a summary of the status of biosimilars around the world. In the plus column, she reports that 14 new biosimilars have been introduced in some countries of the ICH region since the EU issued the guidance in 2005. However, some smaller country markets have been ignored, possibly because of a low potential return on investment.
Naming of a biotherapeutic
Even something as simple as naming the biotherapeutic represents a very high hurdle. A bit of history will explain why this is true. Several years ago, the FDA transferred the responsibility for biotherapeutics from the Center for Biologics Evaluation and Research (CBER) to the Center for Drug Evaluation and Research (CDER). One reason was that CBER operates under the Public Health Service Act, which was focused on vaccines. This act does not support generics. But CDER operates under the Food, Drug and Cosmetic Act, which authorizes generic drugs. Regulations for generics have evolved over time to require an International Nonproprietary Name (INN) and interchangeability of the products. Neither of these fits biosimilars today.
The INN required by CDER and ICH must be different than the brand name of the originator’s product. This is easy in the small-molecule world—for example, DuPont’s Coumadin has an INN of sodium warfarin. The drug active is well known, including its structure and impurity profile. But what is the chemical name for a particular antibody harvested from a particular cell line using a proprietary process? Today’s technology for complex biologics does not support assigning a generic name to an antibody or antibody fragment.
Even if the protein sequence is known, what about post-translational modifications (PTMs) such as glycosylation and phosphorylation? And there are more than 100 other PTMs that are potentially covalently bound. Another complication is that some researchers are seeing evidence that dynamic PTMs also affect higher-order structure. Thus, structure-based naming is not a trivial exercise: If you get it wrong, mistakes will be made and people will suffer. Can we try to fit all of this into the regulatory framework operating well for small molecules?
Interchangeability of complex follow-on biologics
Current CBER and ICH guidelines recognize that materials with the same INN are indistinguishable in safety and efficacy. Thus they are interchangeable between therapeutics from any manufacturer. In practice, interchanging is common. With complex biosimilars, however, few feel that the current state-of-the-art supports interchangeability. Unexpected immunogenicity is the leading fear. Further, it seems dangerous to plan on introducing an immunotherapeutic to the market without clinical testing in humans. Immunogenicity can be a rare event, so backwards traceability supporting postmarket surveillance is required, a difficult feat with today’s INN system.
Is there a solution to this dilemma? What about a multidigit number with several fields for drug type, source, shelf life, etc.? This could be machine readable and facilitate tracing the dose from the factory to the patient and patient record. Emerging semantic technology could establish common and dissimilar points between doses as the information is developed.
Going forward, Dr. Ulm proposed a paradigm for global biosimilars development: a pyramid that starts with physiochemical characterization (ideally of a global reference product) leading upwards to biochemical characterization, preclinical, pharmacokinetics/pharmacodynamics (PK/PD) of reference and candidate products. The apex is comparative clinical trials, but this would be conducted in one jurisdiction only and referenced globally. Such a protocol would be a big step toward global harmonization of the regulatory process.
The problems above are daunting, and so are the technical problems for the followon developer. First, the material in the originator’s biologics license application (BLA) package is proprietary and cannot be divulged by the FDA. In most cases, the prohibition extends to the specifications and methods in the BLA or similar package.
The follow-on developer needs to analyze samples of the reference drug obtained around the world in an attempt to estimate the acceptance criteria for the original product. Without compendial (USP) methods, there is no assurance that the methods used by the two firms will produce comparable results. The follow-on developer needs to use the data obtained from analysis of commercial competitive products to set its own lot release criteria.
That is not so easy, since the originator’s specifications for lot release are probably wider than the range of actual results. No out-of-specification result should ever be commercially distributed. In short, the sampling protocol has a bias. Thus, the follow-on developer would have a tighter specification than the originator, with a corresponding economic disadvantage.
Method transfer is difficult with labs that want to cooperate. There are no incentives for the originator to help the follow-on developer, so cooperation in method transfer is unlikely. This issue is compounded by the fact that samples change with time and analytical methods age as well. It will probably be a decade or two from the time the originator validates a method to when the follow-on starts working on the biosimilars. By then, the methods and instruments used by the originator will have probably been replaced. One vignette at the meeting showed the “advanced analytical instruments” used for lot release of a biotherapeutic in the 1980s. These included slab-gel electrophoresis apparatus and a strip chart recorder. I can imagine the response of the regulators if these played a major role in a new license application.
Another hurdle is the analytics associated with early bioassays, which are used for measuring potency and may be part of lot release. Bioassays for potency may involve obscure endpoints and involve proprietary reagents.
The need for reference standards
It would be helpful if certified analytical reference standards of biotherapeutics were available. They should cover complex proteins with heterogeneity in PTMs and positions. Several scientists from NIST were in the audience. Hopefully they will be able to quickly start to fill the pipeline with useful reference standards.
In the U.S., biosimilars face two unique high hurdles. The Patient Protection and Affordable Care Act (PPAC) prescribes a complicated process with the stated goal of accelerating the approval of biosimilars. It contains provisions requiring interchange of proprietary information. Protecting the IP of both the originator and follow-on products appears very complicated, lengthy, and possibly not generally workable
In early February 2012, the FDA released drafts of three guidance documents for public review. In mid-May, the FDA held a public hearing for oral expression of concerns. Stakeholders seemed skeptical of safety and interchangeability not supported by separate clinical trials. One should expect that these concerns will take more time to resolve, if they can be resolved at all. The lack of approved guidance documents prolongs the uncertainty surrounding biosimilars in the U.S.A.
These difficult hurdles mean the race to follow-on biologics is probably going to be a slow slog, particularly in the U.S., and an even slower process for the more complex products.
Robert L. Stevenson, Ph.D., is a Consultant and Editor of Separation Science for American Laboratory/ Labcompare; e-mail: firstname.lastname@example.org.