EMA Publishes Guidelines for Biosimilar Antibodies - Part I

By James DeGiulio --

On November 23, the European Medicines Agency (EMA) finally published its guidelines for biosimilar monoclonal antibodies (mAb).  The two-part guidelines are now subject to public consultation until May 31, 2011, after which the guidelines are to be finalized.

At first glance, it appears that the EMA has relaxed several classic requirements for equivalence, placing an emphasis on minimizing human and animal subjects whenever possible.  The EMA may also be mindful of testing costs, for an effort is made to minimize repetitive studies by allowing combinations of PK, PD, and safety testing, as well as only requiring analysis of only one clinical indication for each mAb.  The first set of guidelines will be summarized in today's post, and the second set of guidelines covering immunogenicity standards will be discussed in a subsequent post.

The first set of EMA guidelines, entitled "Guideline on similar biological medicinal products containing monoclonal antibodies," is divided into two sections, covering non-clinical and clinical studies.  The non-clinical section begins by advocating a risk-based, case specific approach to evaluate mAbs.  In vitro studies are presented as a preliminary evaluation phase to determine what in vivo clinical testing will be required.  The guidelines spell out a three-step strategy for evaluating the biological activity of mAbs.  The first step is comparing the biosimilar to the reference product via in vitro pharmacodynamic (PD) studies, which must include:  (1) antigen binding, (2) Fc-gamma receptor binding, (3) Fab-associated functions, and (4) Fc-associated functions.  Due to the enhanced sensitivity and specificity of these assays, they must cover all functional aspects of the mAb, even those aspects which are not clinically relevant.

The second step is the identification of factors of importance for the upcoming in vivo non-clinical strategy.  Several factors are presented, including differences in process-related impurities due to cell line choice, potentially uncharacterized impurities, differences in formulations or excipients, need of patient versus volunteer testing, and availability of an animal model for comparative studies.

In vivo studies constitute the third step, which is only authorized if the results from steps 1 and 2 are favorable.  The study should be designed to maximize information, and combinatorial studies on safety, PK, and PD are allowed.  Large toxicology studies on non-human primates are discouraged, despite acknowledgement of the species-specific aspects of mAbs.  Generally, the guidelines relax many of the strict comparative tests seen to this point, and instead stress minimizing the quantity of animal subjects for in vivo studies whenever it is possible.

The next section of the guidelines covers the clinical testing requirements for establishing biosimilarity.  Pharmacokinetic (PK) studies are first detailed, and a parallel group design is acceptable depending on the properties of the mAb.  The guidelines require justification of the population selected for the studies, the goal being maximizing sensitivity and clinical efficacy.  Interestingly, for mAb licensed in several clinical indications, it is sufficient to perform only one PK study.  Multidose PK experiments are detailed as well, suggesting steady state levels for certain types of mAb, noting the long half-life and loading dose interval of mAb.  The guidelines acknowledge that equivalence margins may require some flexibility since many mAbs are highly subject-variable.  Margins beyond the conventional 80-125% will be allowed with sufficient justification.  Rather than require PK profiles done in advance, the guidelines also acknowledge that PK studies may have to be proven clinically due to subject variance.

Regarding pharmacodynamics (PD), as with the animal studies, the guidelines permit combination PK and PD clinical trials to minimize expense, repetition, and human subjects.  Low-dose mAb administration to establish dose-concentration-response relationships between reference product and biosimilar should be done whenever possible.  However, care need be taken not to sensitize human subjects to develop anti-mAb antibodies and make them treatment resistant.

Clinical efficacy studies should only be performed if PD studies do not show comparability in a clinically relevant manner.  The EMA authorizes deviating from the guidelines issued by the Committee for Medicinal Products for Human Use where needed to secure approval, such as use of an interim endpoint for licensing, followed by a post-authorization study with a final endpoint if needed.  This type of deviation is also of particular importance in anticancer mAbs, as typical endpoints such as disease free, progression free, and overall survival may not be sensitive enough.  Various timepoints and strategies are suggested in such a study design.

Clinical safety is the last of the scientific guidelines presented, and use of the same parameters as the reference mAb is recommended if known.  Of course, adverse reactions such as immunogenicity must be addressed in comparative studies.  In most cases, similar pharmacovigilance activities as those of the reference product will be required, rather than a direct comparison with the reference product.  Due to mAb immunogenicity concerns, re-treatment of patients must be addressed by the biosimilar sponsor, and may require a post-authorization study following several treatment cycles.  The EMA immunogenicity guidelines set will be covered in detail in Part II of this series of posts.

Finally, the EMA reiterates that only a single clinical indication need be studied for most biosimilar mAb, unless uncertainty exists, such as with an immunomodulator and a cytotoxic anticancer antibody.  The knowledge in the literature of the different mechanisms of action for the mAb will require evaluation for extrapolation of some clinical indications.

The biosimilar sponsor will be required to present a risk management and pharmacovigilance plan in accordance to EU legislation.  This plan must include post-authorization studies which will evaluate:  (1) the safety of extrapolated clinical indications of the mAb, (2) occurrence of rare and serious adverse effects previously shown by reference product, and (3) detection of novel safety signals.  The EMA admits that this concept may have to exceed routine pharmacovigilance.

The U.S. Food and Drug Administration will likely be paying attention to the EMA's set of guidelines, especially on the heels of the FDA's public hearings on biosimilars earlier this month, where the FDA sought comment on a number of issues surrounding the implementation of the Biologics Price Competition and Innovation Act of 2009 (see "Wrapping up the Issues from the FDA Hearings on Biosimilars").  Certainly apparent at the public hearings was that the standard for establishing biosimilarity and the involvement of clinical trials were issues of particular concern to both patient groups and reference product advocates (see "Clinical Trial Requirements Are Top Issue at FDA Hearings on Biosimilars").

Comments

[DHC]

Can you advise where the EMA provided guidance (if they have) on the nature of the underlying STRUCTURAL identity required in addition to the testing? I didn't see anything that specified sequence homology, glycosylation similarity, or any other structural requirements. Is there any requirement for process similarity (cell type in which the MAbs are synthesized)?

[James DeGiulio]

DHC:

It appears that the EMA would want biosimilar applicants to address those "factors" at the step 2 in vitro testing phase. There are no explicit structural guidelines presented, although cell line choice is mentioned in the context of potential process impurities.

Since each biosimilar mAb situation will present different obstacles, the way I read the guidelines is if a biosimilar applicant seeks to produce a mAb in a different cell line, that will have to be justified to the relevant authority, and additional testing would be required (in this case only) to verify that no harmful impurities are generated from use of the different cell line. But use of another cell line is theoretically permitted as the guidelines currently stand.

As long as the biological activity (and safety profile) of the biosimilar mAb is comparable to the reference product, the structural equivalance of the biosimilar mAb does not appear to be critical. Of course, substantial structural differences between the reference product and the biosimilar would likely alter the biological activity as well.

Thanks for the comment.

James