The Five Key Decisions for Every Biosimilar Development Program
Kurt Brorson
Cecil Nick
Parexel
@Parexel
I

n 2015, FDA approved the first biosimilar drug for the US market. Since then, the agency has approved 33 biosimilars, 21 of which were available commercially as of January 2022. By 4Q 2021, the number of compounds enrolled in FDA’s biosimilar biological product development program had reached 100. The momentum of biosimilar development is building, with development pathways evolving at different rates in different regions. That said, every sponsor will face important decisions along the way that will require careful consideration. This article discusses five of the most critical junctures in biosimilar development.

1. How long should companies monitor the innovator product’s key quality attributes?

It is important to examine many lots of the reference product over an adequate time frame to get a good picture of its quality range. FDA guidance is to examine a minimum of 10 lots. However, key quality attributes in the innovator product can fluctuate considerably over time. There may be periods during which an attribute is at one end of its range and then reverts to the other end for a few months. Capturing too narrow a part of this curve can skew the picture of the product quality range.

Some companies, after sourcing innovator lots over a compressed period, have observed that innovator quality ranges for quality attributes measured by some assays are narrow, even compared to information in the scientific literature. One can propose to FDA that this narrow range may be due to capturing a short window of the innovator product quality profile as it oscillated over its life cycle, but this approach involves extensive background research and carries a substantial risk of the agency requesting additional information and analytics and questioning the company’s interpretation. The agency is bound by the 2009 Biologics Price Competition and Innovation Act to treat each case as unique and cannot use literature-reported innovator product quality ranges as a basis for biosimilarity claims.

Getting as broad a picture of the quality range as possible provides more statistical leeway for the biosimilar product and guards against trapping a company in a narrow band of quality attributes not representative of the actual quality range. A broader picture also relaxes the constraints on manufacturing controls and the target ranges for approval.

A good foundational decision is to purchase batches of product from the innovator over a sustained period. FDA is most concerned with the independence of the drug substance sources; therefore, it is wise to acquire lots with diverse expiry dates, as different drug product batch numbers may not reflect different drug substance sources if their expiry dates are the same. The number of lots is important, as is the period they are collected.

Ideally, lots would be collected over several years, beginning early during research and development. Freezing aliquots is permissible. This way, the sponsors can decide in which direction to develop their product: the earlier, the better.

2. Which manufacturing process should a company use?

Biosimilar IND sponsors must decide whether to replicate what is known of the originator’s manufacturing process or develop a better process. While a sponsor might not know the originator’s exact process, it may know the cell type and other relevant information.

Technology most likely will have advanced since the originator devised its process. Equipment, unit operations, and analytics may all have improved and allow for more efficiency. With price being a considerable commercialization factor, anything that lowers manufacturing costs can make a product more competitive and profitable.

But there is a trade-off: A new process or formulation increases the risk of amplifying quality differences from the originator product. This is particularly true for complex post-translational modifications, such as glycosylation.

Determining the manufacturing process and analytics can be complicated. For the innovators, a protein content method to measure this may have been developed 20 or more years ago. In contrast, biosimilar companies must develop a comparable method, but contemporaneous with their own program. The dilemma for biosimilar firms: how to match the originator’s actual protein content and fulfill the label claim as faithfully as possible if two assays were involved in the product history? A 2021 FDA document clarified that firms should match the protein content of the originator drug using the same assay for both.

3. When should a company engage with the FDA?

It is important to distinguish between the FDA’s Biosimilar Initial Advisory (BIA) and Biosimilar Biological Product Development Type 2 (BPD2) meetings. One key concern is cost: the BIA meeting is free but the BPD2 meeting is not. But this progression can be complex. If the FDA thinks development is not far enough along, it won’t grant a BIA; if it decides development is too far along, it may convert the BIA into a BPD2.

For example: For a BIA meeting, FDA expects a company to have one or a very few pilot batches of its product, just enough to provide preliminary comparative analytical similarity data compared to the US-licensed reference product. However, if development has progressed to the point that a few batches have been created, FDA may convert the meeting to a BPD2. In this case, the company loses the opportunity to discuss the development direction, including fundamentals that might keep the product from being approved. Companies should take care not to let a program inadvertently advance beyond the initial concept stages suitable for a BIA because it’s better to learn some things at a BIA meeting rather than further down the development stream.

4. What clinical data should a company collect?

In 20 years of biosimilar development, assay technology has significantly improved, and our understanding of the impact of variants on efficacy and safety has, too. Today, physicochemical and biological testing coupled with comprehensive pharmacokinetic (PK) data can provide strong evidence of biosimilarity and has brought into question the need for equivalent efficacy trials in patients.

A 2020 study of 42 biosimilar programs found that not one was rejected because of differences in efficacy. Where differences in therapeutic response have been observed, they have been explained by prior immunogenicity, protein content, or structural differences. Consequently, the UK’s Medicines and Healthcare Products Regulatory Agency (MHRA) has stated that, generally, only a pivotal PK study is required to show clinical comparability, not a comparative efficacy trial. However, confirmation of this approach rests on a final assessment of the total data package.

Both FDA and the European Medicines Agency (EMA) require that in the absence of surrogate markers, a comparative efficacy study must be undertaken. A literature search can help to identify reference data to statistically justify a broad equivalence margin and thereby reduce the sample size, though the margin must also be clinically justified. Trial design must accommodate differences in approved indications, dosage, trial participant characteristics and demographics, and agency requirements among regions (for a single global study). Its locations will be influenced by where the treated condition is most prevalent and where the reference drug is less readily available.

5. Should interchangeability be a goal?

EMA considers an approved biosimilar interchangeable. It can be used instead of its reference product or another biosimilar of the same reference product, but individual EU member states regulate the role of the prescriber in interchangeability.

In the US, interchangeable status allows the substitution of a biosimilar at the pharmacy level. To date, the FDA has granted interchangeability status to three biosimilars: Semglee, Cyltezo, and Cimerli. FDA did not consider the clinical multiple switch study submitted by the sponsor for Semglee to be adequate but granted interchangeability based on high similarities in its quality attributes, which FDA considered obviated the need for a clinical immunogenicity study. Cyltezo was first approved as a biosimilar in 2017 and won interchangeability status in 2021, though it cannot be marketed until 2023, when certain patents expire. Cimerli, like Semglee, was granted interchangeability status based on quality similarity without the need for additional clinical data. Several other companies have announced plans to pursue interchangeable status for their biosimilars.

Interchangeability status is of most value in primary care. For hospital-based products, physicians are on-site to approve whatever the hospital dispenses. Interchangeability approval has typically required a 12-month, approximately 400-subject trial and was often not considered worth the expense in regulatory jurisdictions where physicians could easily approve a substitution. Since two of the three approved interchangeable biosimilars were approved based on quality similarity and without the need for additional clinical data, this view could shift. However, Cimerli and Semglee are simple non-glycosylated proteins for which the clinical approach recommended in the FDA guidelines would not be feasible. It remains to be seen what requirements might be imposed on future interchangeable biosimilar candidates.

Getting Ahead of the Game

These five major considerations apply to every biosimilar development journey. They may not provide every answer, but they have two things in common. First, the sooner they are confronted, the more time will be available for executing them. Second, they should be made consciously and not by default. Addressing them in a timely and deliberate fashion will lead to a shorter and smoother development and approval journey.
References available upon request.