Regulators See Trend To Reduced Clinical Data Requirements For Biosimilars

The amount of clinical trial data needed to support a biosimilar drug application is declining as more experience is gained with biosimilar use and more use is made of increasingly sensitive analytical tools. Moreover, the concept of extrapolation of indications is gaining more credence, and the idea of immunogenicity as a particular problem with biosimilars seems to be losing traction.

These were among the views put forward at the recent conference organized in London by Medicines For Europe's Biosimilar Medicines Group¹, where speakers looked at issues regarding biosimilar preclinical, clinical and quality data requirements, extrapolation and immunogenicity, and examined how far the EU regulatory requirements have evolved in the decade since Sandoz' growth hormone Omnitrope (somatropin) became the first biosimilar to be approved in April 2006.

The EU was the first jurisdiction to introduce a legal framework and guidance for biosimilars, and initially the concept was science driven and stakeholders lacked practical experience, so a "cautious, conservative approach" was taken, said Martina Weise of the German regulatory agency BfArM, who is also vice-chair of the European Medicines Agency's biosimilar medicines working party (BMWP).

Initially there were doubts about the suitability and sensitivity of the existing analytical methods to adequately characterize biosimilars, especially versions of complex biologic drugs, Weise noted. Moreover, a non-clinical in vivo toxicology study was required by default, as were clinical safety and efficacy trials.

But after "10 years of safe and effective use of biosimilars," she said, the EU is moving from a purely science-driven conceptual approach to a science-driven knowledge-based approach, with a growing armamentarium of increasingly sensitive analytical tools that allow the clinical drug development program for biosimilars to be tailored and reduced.

This has taken place through the development and revision of the biosimilar guidelines on the basis of advances in analytical sciences, and lessons from post-approval manufacturing process changes, biosimilar product reviews, and scientific advice procedures.

Interactions with stakeholders have also played a key role, Weise said, via the International Pharmaceutical Regulators Forum's biosimilars working group, the World Health Organization, and other bodies, as well as external consultations on draft guidelines, EMA/European Commission workshops, and participation in learned society meetings.

Looking at the key changes in requirements for biosimilar approval over the past 10 years, she said that for non-clinical studies the focus was now "very much on comparative in vitro studies to extensively characterize and compare the functions and activity of the molecule," which are usually more sensitive to detect differences than in vivo studies. There is also an increasing reliance on taking a risk-based approach to in vivo studies, for example requiring them in cases where a novel expression system or excipients are involved or the quality findings are of concern or unclear relevance.

In the area of quality data, there are no "factual" changes to the requirements, Weise said, but the guidance on what data are needed has been made clearer, particularly on issues like employing state-of-the-art analytical tools and the need to identify the "quality target product profile" based on characterization of multiple batches of the reference product.

She also noted that global development of biosimilars is more streamlined now that companies can compare a biosimilar to a reference product that is not authorized in Europe as long as it is approved using similar scientific and regulatory standards to those of the EMA, and the drug is representative of the reference product in the EU. In such cases, she noted, bridging data are needed, including a three-way analytical comparison that may include a three-way pharmacokinetic (PK) and/or pharmacodynamic (PD) comparison.

Clinical Data Requirements

But perhaps the biggest change is in the clinical data requirements for biosimilars. Weise said the EMA was "more and more comfortable relying on analytical tools," and that this would "finally lead to a clear reduction in the biosimilar clinical development program."

Noting that clinical studies with biosimilars are not intended to show efficacy and safety per se, she said that the endpoints used should nonetheless be sensitive enough to detect differences in efficacy, and may be different from those used in clinical trials of the originator.

"There is a clear shift to pharmacodynamic endpoints" such as MRI data for multiple sclerosis drugs, overall response rate for oncology medicines, absolute neutrophil count for G-CSF biosimilars, and glucose infusion rate in insulin clamp studies, Weise noted.

Usually one clinical trial will be needed, with the option to extrapolate. However, a study may not be required if the analytical, PK and PD comparisons allow reviewers to conclude that the biosimilar has similar efficacy and safety and that the impurity profile and nature of the excipients do not give rise to any concerns, Weise continued.

This approach, she said, is mentioned in the general EMA guidelines, specifically for insulin and LMW heparin, and is planned for filgrastim. It is also expected to apply to more complex biologicals as it becomes easier to characterize their physicochemical and functional properties.

Extrapolation too is becoming more of an established concept, and regulators consider that it should now be seen "in the light of the totality of the data supporting biosimilarity," Weise said. The focus now is less on situations where extrapolation "may be considered" and more on "situations were additional data may be required."

In the discussion that followed, Peter Richardson, head of quality at the EMA, agreed that after a "slightly conservative approach" at the beginning, "we are now getting to the meat of what we can do with biosimilars and we will see how much we can reduce the clinical trial requirements. We have gone a long way but we still need a little bit of caution."

Immunogenicity

One key topic raised was the question of immunogenicity of biosimilars and how far this needs to be addressed in the comparability exercise. One delegate suggested the immunogenicity problem had perhaps been "overestimated," noting that post-approval studies had shown no instances of biosimilars inducing unique immunogenicity events.

Weise said that at present "we can't fully rely on the analytical comparison to conclude on immunogenicity," although she added that the trend was probably in that direction. Christian Schneider, chair of the BMWP, suggested the regulators might be open to more preclinical in vivo studies to investigate immunogenicity, and that some companies were heading this way, although "very in-depth expertise" would be required.

Others were less cautious. Noting that regulators do not ask for immunogenicity data after changes have been made to the manufacturing process for originator biologics, Frits Lekkerkerker of consultancy company NDA Regulatory Sciences said immunogenicity was a real issue but that the clinical relevance of the appearance of antibodies to a biosimilar was not always clear. "A very sensitive assay can always find antibodies. Maybe we over-value these assays and should look more at the kinetic data," he suggested.

References

1. 14th Annual Biosimilar Medicines Group Conference, London, April 28-29