Biosimilar Series Part I: Overcoming statistical challenges in biosimilar trials

With the onset of patent expirations for several biological products, there is a steady rise in the development of alternate versions, termed "biosimilars" which have comparable quality, safety, and efficacy to the innovator biologic. As the name suggests, they are "similar” but not exact replicas of the innovator product. Thus, the main objective of clinical trials for biosimilars is to demonstrate the similarity between the biosimilar and the innovator for efficacy, safety, and immunogenicity. In addition, these trials also establish that there are “no clinically meaningful differences” between the two.?

Statistical challenges encountered:

Equivalence vs superiority

Traditional trials typically concentrate on how well a new treatment outperforms the currently accepted standard of treatment. Biosimilar trials, however, attempt to demonstrate equivalence or non-inferiority. As a result, the objectives require different statistical techniques such as equivalence and non-inferiority testing.

Variability

Biosimilars are large, complex molecules made inside living cells. The inherent micro-variability is one of the main challenges. Even small variations in the manufacturing methods of the biosimilar versus the innovator could have a huge impact on the final product. Typically, biologics trigger an immune response, leading to the production of antibodies. This is called immunogenicity, which is extremely variable and can impact the safety and efficacy of the drug. Navigating the multitude of variations and sources makes the design and interpretation of clinical studies quite challenging.

Sample size determination

Biosimilar studies typically involve a large number of participants, to meet the study objectives. An adaptive design approach allows for the trial strategy to be redefined based on the initial study, thus making the trial smaller and reducing exposure. A Blinded Sample Size Re-estimation (BSSR) is another good technique to re-evaluate the sample size based on a blinded review of the data, to derive necessary insights.?

Another variable that has an impact on sample size estimation is the equivalence margin. The smaller the equivalence margin, the larger the sample size required to demonstrate equivalence and vice versa. If the range is too narrow, it will lead to a complex, lengthy and costly trial. On the other hand, if the margin is too wide, it may not be clinically meaningful. Hence, determining an appropriate equivalence margin plays a critical role in designing biosimilar trials.????

Multiple comparisons

When dealing with multiple endpoints such as efficacy, safety, immunogenicity, and PK/PD, Type 1 errors will be on the rise. This could lead to incorrect trial conclusions due to multiple comparisons. The overall type 1 error rate must be controlled with statistical adjustments, to ensure the trial isn’t too large, time-taking, and expensive.

Interchangeability

The trial design might need to incorporate a switching study where patients alternate between the biosimilar and the innovator to prove interchangeability. Compared to a straightforward comparison study, this study design is more complex and involves more participants and time. Additionally, the rules for establishing interchangeability vary between regulators. For example, unlike the FDA, the EMA (European Medicines Agency) doesn't account for interchangeability in its guidelines, making the regulatory landscape challenging to navigate.?

Also, when participants are switched back and forth between the innovator and the biosimilar, there may be concerns about changes in efficacy and the emergence of adverse events, including immunogenicity. Thus, interchangeability adds a layer of complexity to the already challenging process of designing biosimilar trials.?

In conclusion, designing biosimilar trials require careful consideration of all potential statistical challenges to demonstrate similarity and equivalence to the reference product. Addressing these early on is crucial for generating reliable clinical outcomes, smoother regulatory approvals, and easier acceptance by healthcare professionals and patients. As the demand for biosimilars continues to grow, extensive collaboration among statisticians, regulators, and industry experts is essential to making biosimilar clinical development simpler and more efficient.?

To reach out to Algorics Biostatistics Centre of Excellence, write to us at [email protected]

?