More Than a Decade After the First Biosimilar, the Regulatory Landscape Is Increasingly Complex

Stakeholders are well aware that biosimilar developers are increasingly considering their development programs to be global in scope, but challenges remain in meeting the demands of regulatory bodies in different jurisdictions. During the first day of the BioTech Pharma third annual Biosimilars and Biologics Summit, held March 21 to 22 in Porto, Portugal, Barbara Valenta-Singer, MD, chief medical officer of biosimilars at Fresenius-Kabi SwissBioSim, gave an expert perspective on the past 10-plus years of experience with biosimilars in the European and US contexts.

Valenta-Singer began her presentation with a discussion of the first biosimilar approved in the European Union: Omnitrope, a biosimilar somatropin that gained European Medicines Agency (EMA) clearance in 2006. In the United States, the drug was approved 2 years later, though there was no biosimilar approval pathway in the United States at the time, and the drug is therefore considered a follow-on in the United States rather than a biosimilar.

Valenta-Singer explained that the data package for Omnitrope included 3 pharmacokinetic (PK) studies in healthy volunteers, and a single pivotal study in just 51 patients. “At the time,” she said, “the appetite to get something new was very high. The threshold of scrutiny, the concern, was much lower.”

That early experience stands in stark contrast to products approved in 2018, in which large PK and pharmacodynamic studies were undertaken, and “In oncology, we’re talking about 500-, 700-, 800-patient studies” as culminating elements of the clinical programs for biosimilar monoclonal antibodies.

Among the challenges that have emerged and developed over the past decade is the question of reference products used in development programs. By 2018, said Valenta-Singer, only 2 biosimilars had been approved by the FDA on the basis of studies that used US-licensed reference products: Boehringer Ingelheim’s adalimumab biosimilar (Cyltezo), and Celltrion’s trastuzumab biosimilar (Herzuma). The reason? The cost of the US-licensed reference product is substantially higher than the cost of the EU-licensed product.

Using an EU-licensed reference results in the need for a bridging study to demonstrate the equivalence of the EU- and US-licensed originators, a situation that puts additional burden on the biosimilar developer. Valenta-Singer asked, “Why do I need to prove the originals’ bioequivalence? Is this where we want to be?”

The use of bridging studies also has implications for interchangeability as viewed by the United States. In the US context, a demonstration of interchangeability requires multiple switches between the reference and the biosimilar; a developer would likely need to use the US reference as part of such a clinical program, so data that rely on a bridge to the US reference would not be useful in this context, and a new, expensive, time-consuming study might be necessary.

Given that some monoclonal antibodies have long half-lives, and some may be administered only every 6 months, said Valenta-Singer, a demonstration of interchangeability involving multiple switches could be prohibitively lengthy.

Other issues that have emerged for global development of biosimilars include divergent requirements between the FDA and the EMA; whereas the FDA uses 90% standard deviations, the EMA uses 95%. The EMA recommends population PK studies, while the FDA has not expressed a need for such studies that increase the burden on the developer. Whereas the FDA recommends switching studies, the EMA does not require them. Expectations regarding end points may also differ. Finally, the FDA has not provided clarity in terms of the statistical approaches to evaluate analytical similarity after, in June 2018, it withdrew its guidance on the topic and has not issued new guidance. “On the one hand. It gives you more flexibility…on the other hand, it doesn’t really help you if you don’t know,” she said.

There is also some lack of clarity related to how regulators are addressing data packages for biosimilars. Valenta-Singer presented a case study of 4 approved biosimilar trastuzumabs: Ontruzant (approved in the United States and European Union), Herzuma (approved in the United States and European Union), Kanjinti (approved in the European Union), and Trazimera (approved in the United States and European Union). While these biosimilars were all tested in single-dose, parallel-group PK studies in between 105 and 157 healthy male volunteers, the pivotal studies conducted for these drugs differed in some significant regards.

While Ontruzant, Herzuma, and Kanjinti all had difference in complete response rate as their primary end points in their phase 3 trials, Trazimera’s primary phase 3 end point was the risk ratio of the overall response rate. Furthermore, while Herzuma and Trazimera demonstrated equivalence based on their prespecified equivalence margins, Ontruzant and Kanjinti did not demonstrate equivalence.

Despite failing to meet their primary end points in the phase 3 studies, these 2 products were approved on the basis of healthy volunteer PK studies, Valenta-Singer said. This fact raises an important question, Valenta-Singer said: “Why did you do the study if you failed the study and then were approved?” While she acknowledged that it is not known what supporting data are in the analytical package, “There’s still an opportunity if you fail. This is the good message.”

Moving forward, greater regulatory clarity will be crucial for biosimilar developers to gain the certainty that they need in order to continue to undertake these time-consuming and costly development programs for biosimilars intended for a global marketplace.