The Clinical Side of Biosimilar Development

View PDF


Biosimilars have become common on pharmacy shelves in Europe. The first biosimilar product — Sandoz’s Omnitrope version of Lilly’s Humatrope (somatropin) — was approved by the European Medicines Agency (EMA) in 2006. In the decade that followed, more than 20 biosimilars have gained regulatory approval in Europe. The first biosimilar monoclonal antibodies (MAbs) — comparators to Janssen’s Remicade (infliximab) — were approved in 2013.

The pace of approvals in the United States has been much slower. The US Food and Drug Administration (FDA) agreed upon a 351(k) pathway for biosimilar approval in 2010 and published draft guidance documents two years later. Yet it was not until 2015 that Sandoz’s Zarxio biosimilar comparator to Amgen’s Neupogen (filgrastim) got the go-ahead for US marketing. Multiple filgrastim biosimilars (including the Sandoz product under the brand name Zarzio) already had been licensed in Europe before the US approval.

Since then, the number of approvals has rapidly increased, with several biosimilar products now on the US market. Some big pharmaceutical companies have made substantial investments in developing further biosimilars, and a number of marketing authorization applications are awaiting consideration by the FDA now.

In essence, biosimilars could be seen as a “generic” biologic products. However, their approval process is somewhat more complex than it is for small-molecule generics. A biosimilar will offer cost savings to patients relative to an originator product; in practice, that price differential is unlikely to be large. The path to market for biosimilars involves an abbreviated approval process focused on proving their “equivalence” to reference originator products.

A clue to the other significant difference is in the word biosimilar. The active ingredient in a small-molecule generic necessarily is exactly the same as an originator. With biologics, that is not the case. Competitor products are almost the same, but not exactly identical — hence the similar. Inevitably, slight differences occur in their structure as a result of different manufacturing processes, and it is the responsibility of each biosimilar producer to prove that those structural differences have no influence on a product’s safety and efficacy relative to the originator.

Generic substitution is mandated in the United States, where pharmacists will dispense whichever brand of a small-molecule generic they happen to have in stock at any given time. However, biosimilars are different. Additional studies are required to prove that such products are “interchangeable” with their comparators, which requires a carefully formulated clinical development program.

Another important consideration is selection of the indication for which approval will be sought. Many biologics (particularly MAbs) are marketed for a number of different indications. So developers must decide at the outset which offers the best route to initial approval. If biosimilarity is demonstrated for one indication, then in some cases similarity could be extrapolated to other indications for which the originator product is marketed — with appropriate scientific justification.

The All-Important CMC Dossier
With all pharmaceutical products, the chemistry, manufacturing, and controls (CMC) dossier is an essential part of the regulatory submission package for a clinical trial — and in later stages, an application for market authorization. The CMC part is even more critical to biosimilars, for which the dossier must include a head-to-head comparison against the originator product. As part of a comprehensive physicochemical and biological characterization meant to establish the impact of all differences on a product’s safety and efficacy, a biosimilar’s CMC dossier points out where the points of difference between the two products are and how they affect the product. It will include details of analytical and other methods used to identify those differences. And that will allow a regulatory assessor to decide just how similar the two products actually are. That comparison is the core of a biosimilar’s CMC dossier.

It also needs to contain details about how a product is manufactured, describing cell lines used and the sources of all materials that accompanied them in the bioreactor. The dossier also describes all process control methods used and provides information about how analytical methods and data have been validated. All that information is contained in every CMC dossier, whether for a small molecule or a biologic, a generic or a biosimilar.

But head-to-head comparison data are an additional requirement for biosimilars, and producing them is not a trivial endeavor. Such data are rarely available for originator products, so the entire analytical exercise must be performed twice — once on the biosimilar and again on the originator product — to enable that comparison to be made. This is a relatively costly process, and locating a source of an originator product can prove extremely difficult.

A comprehensive set of preclinical safety studies also must be carried out before human volunteers or patients are dosed with a potential biosimilar. These will include in vitro assays and appropriate animal models, which are designed to predict whether those small differences may affect safety or efficacy.

Immunogenicity is a particular concern. Both in silico tools and in vitro assessments using animal tissue can be used to predict whether it is likely to be a problem in human patients. Animal immunogenicity studies also must be carried out before humans are dosed for the first time.

Biosimilar Phase 1 Studies
For an originator product, phase 1 safety studies typically involve 30–48 healthy volunteers. By contrast, the study group for a biosimilar phase 1 study will be made up of 120–200 healthy volunteers. That discrepancy arises not because the trial is intended to establish primary safety; rather, it is designed to detect differences in safety signals between the biosimilar and its originator comparator.

Although structural differences between the two products are necessarily small, they do introduce potential risk, even though analytical methods and preclinical animal studies could suggest that human patients should experience no differences. A biosimilar phase 1 study must show that the product is safe, but a developer also must prove that it is just as safe as the original. That requires statistical calculations, and a larger group of subjects is required for those calculations to be meaningful.

Some countries are better choices than others for locating such studies. It is important to select one where the local health authorities and ethics committees understand what a phase 1 biosimilar study’s aims are, how they differ from innovator phase 1 safety studies, and why so many subjects are required.

Each trial subject will receive either the originator product (its safety already established) or biosimilar (with uncertain safety) dose. So there is a small but definite chance that those seemingly insignificant structural differences — or other minuscule differences that were not detected in analytical/preclinical work — could cause a major adverse event such as an immune cascade. It could be argued whether it is ethically acceptable to expose healthy volunteers to such risks just to test a product that is essentially a copy of something that is already on the market (and of which the added medical value might be debated).

That said, the best place to operate a study of this nature is in a country that already has a good phase 1 infrastructure and a favorable regulatory environment, where study-approval timelines are not too long. Biosimilar phase 1 trials commonly are carried out in the United Kingdom, Belgium, the Netherlands, and/or the United States. They are seldom run in Germany or France, which seem to be less accepting of biosimilar studies.

At the outset, it is important to have discussions with relevant regulatory authorities, whether that be the EMA, FDA, British Medicines and Healthcare Products Regulatory Agency (MHRA), or another authority in a different part of the world. Such meetings are meant to establish acceptability of the study design and whether (based on the CMC dossier) a product may be considered by the agency to be biologically similar to the originator in light of the documented structural differences between the two. If regulators do not determine the products to be potentially biosimilar, there is little point in continuing with its development as such because it is unlikely to gain marketing authorization by that biosimilar pathway. (“Biobetter” filing may be an option.) It is vital to engage in these discussions with regulators very early on to ensure that precious time, money, and resources are not wasted.

Broadly speaking, the planning process with regulators for a biosimilar market authorization application are similar around the world. EU guidelines are further advanced than those in the United States, however, having been implemented for a longer period.

In addition to the phase 1 comparative safety study, a phase 1b study is required to prove equivalent efficacy. Again, it is important to discuss the design of this study with regulators and together establish what endpoints will be required. Such trials need to involve representative patient populations with sufficient numbers of subjects to ensure that it will detect potential differences in efficacy between the two products.

In the United States, the additional aspect of interchangeability must be established. In the European Union, important questions of pricing and reimbursement must be addressed. As with all applications for drug approval, discussions with regulators are necessary to review the inevitable data differences that will be observed and how best to manage them.

If deviations are made from the reference product — e.g., strength, pharmaceutical form, or formulation — they will need to be justified. However, it is important to note that a biosimilar’s route of administration must be the same as the reference product. Some changes will be incompatible with biosimilarity — if they change the molecule’s chemical nature too much relative to the reference product. An example would be optimizing the glycosylation pattern to improve efficacy (thus creating a “biobetter” product). Changes designed to improve safety may not preclude a biosimilar decision, however; they could include reducing immunogenicity or lowering levels of known impurities present relative to the reference product.

Case Study Examples
As an example of the difficulty in gaining approval to carry out a phase 1 biosimilar study, one large generics company was about to begin such a trial for a biosimilar MAb. However, the proposed trial was rejected in the Netherlands. Key to gaining the go-ahead to run this trial was twofold: completely reworking the CMC dossier and applying to run the trial in another country. Changes to the dossier used already-available information but provided an additional head-to-head comparison of analytical data. The results were submitted to the health authority in Belgium, which gave the trial the green light within 15 days.

Different problems were faced by a Middle Eastern company planning to enter the EU market with a biosimilar MAb. The manufacturing site was located in a noninspectable country, and concerns were raised about whether the product was sufficiently close structurally to its reference comparator. Also questions were raised about whether available clinical data had been collected according to good clinical practices (GCPs).

In that case, the first step toward study approval was to engage in discussions with the EMA and national authorities on the GMP inspection of the manufacturing site. All available data were reviewed, including analytical results on both the biosimilar and reference products, with both preclinical and clinical data. Based on review of all available information, it was decided that the MAb was unlikely to have been considered as a biosimilar.

Those two examples highlight the importance of taking the utmost care in preparation of CMC dossiers. Including as much relevant detail as possible will increase the likelihood that regulators and health authorities will take a positive view of it. As more biosimilars reach the market, regulators are sure to become more experienced and comfortable with the concept of a biosimilar phase 1 study. However, convincing them that the study won’t put volunteers at risk will remain an essential part of the process. Clear, careful documentation of all analytical and preclinical studies that have been carried out will go a long way toward providing that reassurance.

Bruno Speder is head of clinical regulatory affairs at SGS,;